Spaces:
Running
on
Zero
Running
on
Zero
guangkaixu
commited on
Commit
•
a123370
1
Parent(s):
12daec9
upload
Browse files- app.py +3 -3
- pipeline_genpercept.py +1 -5
app.py
CHANGED
@@ -79,8 +79,8 @@ def process_image(
|
|
79 |
show_progress_bar=False,
|
80 |
)
|
81 |
|
82 |
-
depth_pred = pipe_out.
|
83 |
-
depth_colored = pipe_out.
|
84 |
depth_16bit = (depth_pred * 65535.0).astype(np.uint16)
|
85 |
|
86 |
np.save(path_out_fp32, depth_pred)
|
@@ -266,7 +266,7 @@ def main():
|
|
266 |
|
267 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
268 |
|
269 |
-
dtype = torch.
|
270 |
|
271 |
vae = AutoencoderKL.from_pretrained("guangkaixu/GenPercept", subfolder='vae').to(dtype)
|
272 |
unet_depth_v1 = UNet2DConditionModel.from_pretrained('guangkaixu/GenPercept', subfolder="unet_depth_v1").to(dtype)
|
|
|
79 |
show_progress_bar=False,
|
80 |
)
|
81 |
|
82 |
+
depth_pred = pipe_out.pred_np
|
83 |
+
depth_colored = pipe_out.pred_colored
|
84 |
depth_16bit = (depth_pred * 65535.0).astype(np.uint16)
|
85 |
|
86 |
np.save(path_out_fp32, depth_pred)
|
|
|
266 |
|
267 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
268 |
|
269 |
+
dtype = torch.float32
|
270 |
|
271 |
vae = AutoencoderKL.from_pretrained("guangkaixu/GenPercept", subfolder='vae').to(dtype)
|
272 |
unet_depth_v1 = UNet2DConditionModel.from_pretrained('guangkaixu/GenPercept', subfolder="unet_depth_v1").to(dtype)
|
pipeline_genpercept.py
CHANGED
@@ -148,14 +148,10 @@ class GenPerceptPipeline(DiffusionPipeline):
|
|
148 |
# Normalize rgb values
|
149 |
rgb = np.transpose(image, (2, 0, 1)) # [H, W, rgb] -> [rgb, H, W]
|
150 |
rgb_norm = rgb / 255.0 * 2.0 - 1.0
|
151 |
-
rgb_norm = torch.from_numpy(rgb_norm).to(self.dtype)
|
152 |
rgb_norm = rgb_norm[None].to(device)
|
153 |
assert rgb_norm.min() >= -1.0 and rgb_norm.max() <= 1.0
|
154 |
bs_imgs = 1
|
155 |
-
|
156 |
-
print('rgb_norm :', rgb_norm.dtype)
|
157 |
-
print('unet :', self.unet.dtype)
|
158 |
-
print('vae :', self.vae.dtype)
|
159 |
|
160 |
# ----------------- Predicting depth -----------------
|
161 |
|
|
|
148 |
# Normalize rgb values
|
149 |
rgb = np.transpose(image, (2, 0, 1)) # [H, W, rgb] -> [rgb, H, W]
|
150 |
rgb_norm = rgb / 255.0 * 2.0 - 1.0
|
151 |
+
rgb_norm = torch.from_numpy(rgb_norm).to(self.unet.dtype)
|
152 |
rgb_norm = rgb_norm[None].to(device)
|
153 |
assert rgb_norm.min() >= -1.0 and rgb_norm.max() <= 1.0
|
154 |
bs_imgs = 1
|
|
|
|
|
|
|
|
|
155 |
|
156 |
# ----------------- Predicting depth -----------------
|
157 |
|