Spaces:
Running
on
Zero
Running
on
Zero
wondervictor
commited on
Update model.py
Browse files
model.py
CHANGED
@@ -87,7 +87,7 @@ class Model:
|
|
87 |
def load_t5(self):
|
88 |
precision = torch.bfloat16
|
89 |
t5_model = T5Embedder(
|
90 |
-
device=
|
91 |
local_cache=True,
|
92 |
cache_dir='checkpoints/flan-t5-xl',
|
93 |
dir_or_name='flan-t5-xl',
|
@@ -114,8 +114,8 @@ class Model:
|
|
114 |
image = resize_image_to_16_multiple(image, 'canny')
|
115 |
W, H = image.size
|
116 |
print(W, H)
|
117 |
-
self.t5_model.model.to(self.device)
|
118 |
-
self.gpt_model_canny.to(self.device)
|
119 |
|
120 |
condition_img = self.get_control_canny(np.array(image), low_threshold,
|
121 |
high_threshold)
|
@@ -191,10 +191,10 @@ class Model:
|
|
191 |
image = resize_image_to_16_multiple(image, 'depth')
|
192 |
W, H = image.size
|
193 |
print(W, H)
|
194 |
-
self.t5_model.model.to(self.device)
|
195 |
-
self.gpt_model_depth.to(self.device)
|
196 |
-
self.get_control_depth.model.to(self.device)
|
197 |
-
self.vq_model.to(self.device)
|
198 |
image_tensor = torch.from_numpy(np.array(image)).to(self.device)
|
199 |
condition_img = torch.from_numpy(
|
200 |
self.get_control_depth(image_tensor)).unsqueeze(0)
|
|
|
87 |
def load_t5(self):
|
88 |
precision = torch.bfloat16
|
89 |
t5_model = T5Embedder(
|
90 |
+
device=self.device,
|
91 |
local_cache=True,
|
92 |
cache_dir='checkpoints/flan-t5-xl',
|
93 |
dir_or_name='flan-t5-xl',
|
|
|
114 |
image = resize_image_to_16_multiple(image, 'canny')
|
115 |
W, H = image.size
|
116 |
print(W, H)
|
117 |
+
# self.t5_model.model.to(self.device)
|
118 |
+
# self.gpt_model_canny.to(self.device)
|
119 |
|
120 |
condition_img = self.get_control_canny(np.array(image), low_threshold,
|
121 |
high_threshold)
|
|
|
191 |
image = resize_image_to_16_multiple(image, 'depth')
|
192 |
W, H = image.size
|
193 |
print(W, H)
|
194 |
+
# self.t5_model.model.to(self.device)
|
195 |
+
# self.gpt_model_depth.to(self.device)
|
196 |
+
# self.get_control_depth.model.to(self.device)
|
197 |
+
# self.vq_model.to(self.device)
|
198 |
image_tensor = torch.from_numpy(np.array(image)).to(self.device)
|
199 |
condition_img = torch.from_numpy(
|
200 |
self.get_control_depth(image_tensor)).unsqueeze(0)
|