Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,4 +1,5 @@
|
|
1 |
import torch
|
|
|
2 |
import gradio as gr
|
3 |
from main import setup, execute_task
|
4 |
from arguments import parse_args
|
@@ -51,6 +52,7 @@ def clean_dir(save_dir):
|
|
51 |
|
52 |
def start_over(gallery_state, loaded_model_setup):
|
53 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
54 |
if gallery_state is not None:
|
55 |
gallery_state = None
|
56 |
if loaded_model_setup is not None:
|
@@ -63,6 +65,7 @@ def setup_model(prompt, model, seed, num_iterations, enable_hps, hps_w, enable_i
|
|
63 |
|
64 |
"""Clear CUDA memory before starting the training."""
|
65 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
66 |
|
67 |
# Set up arguments
|
68 |
args = parse_args()
|
@@ -108,6 +111,7 @@ def setup_model(prompt, model, seed, num_iterations, enable_hps, hps_w, enable_i
|
|
108 |
|
109 |
def generate_image(setup_args, num_iterations):
|
110 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
111 |
|
112 |
args = setup_args[0]
|
113 |
trainer = setup_args[1]
|
@@ -125,6 +129,7 @@ def generate_image(setup_args, num_iterations):
|
|
125 |
|
126 |
try:
|
127 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
128 |
steps_completed = []
|
129 |
result_container = {"best_image": None, "total_init_rewards": None, "total_best_rewards": None}
|
130 |
error_status = {"error_occurred": False} # Shared dictionary to track error status
|
@@ -175,6 +180,7 @@ def generate_image(setup_args, num_iterations):
|
|
175 |
|
176 |
if error_status["error_occurred"]:
|
177 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
178 |
yield (None, "CUDA out of memory. Please reduce your batch size or image resolution.", None)
|
179 |
else:
|
180 |
main_thread.join() # Ensure thread completion
|
@@ -182,13 +188,16 @@ def generate_image(setup_args, num_iterations):
|
|
182 |
if os.path.exists(final_image_path):
|
183 |
iter_images = list_iter_images(save_dir)
|
184 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
185 |
time.sleep(0.5)
|
186 |
yield (final_image_path, f"Final image saved at {final_image_path}", iter_images)
|
187 |
else:
|
188 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
189 |
yield (None, "Image generation completed, but no final image was found.", None)
|
190 |
|
191 |
torch.cuda.empty_cache() # Free up cached memory
|
|
|
192 |
|
193 |
except torch.cuda.OutOfMemoryError as e:
|
194 |
print(f"Global CUDA Out of Memory Error: {e}")
|
|
|
1 |
import torch
|
2 |
+
import gc
|
3 |
import gradio as gr
|
4 |
from main import setup, execute_task
|
5 |
from arguments import parse_args
|
|
|
52 |
|
53 |
def start_over(gallery_state, loaded_model_setup):
|
54 |
torch.cuda.empty_cache() # Free up cached memory
|
55 |
+
gc.collect()
|
56 |
if gallery_state is not None:
|
57 |
gallery_state = None
|
58 |
if loaded_model_setup is not None:
|
|
|
65 |
|
66 |
"""Clear CUDA memory before starting the training."""
|
67 |
torch.cuda.empty_cache() # Free up cached memory
|
68 |
+
gc.collect()
|
69 |
|
70 |
# Set up arguments
|
71 |
args = parse_args()
|
|
|
111 |
|
112 |
def generate_image(setup_args, num_iterations):
|
113 |
torch.cuda.empty_cache() # Free up cached memory
|
114 |
+
gc.collect()
|
115 |
|
116 |
args = setup_args[0]
|
117 |
trainer = setup_args[1]
|
|
|
129 |
|
130 |
try:
|
131 |
torch.cuda.empty_cache() # Free up cached memory
|
132 |
+
gc.collect()
|
133 |
steps_completed = []
|
134 |
result_container = {"best_image": None, "total_init_rewards": None, "total_best_rewards": None}
|
135 |
error_status = {"error_occurred": False} # Shared dictionary to track error status
|
|
|
180 |
|
181 |
if error_status["error_occurred"]:
|
182 |
torch.cuda.empty_cache() # Free up cached memory
|
183 |
+
gc.collect()
|
184 |
yield (None, "CUDA out of memory. Please reduce your batch size or image resolution.", None)
|
185 |
else:
|
186 |
main_thread.join() # Ensure thread completion
|
|
|
188 |
if os.path.exists(final_image_path):
|
189 |
iter_images = list_iter_images(save_dir)
|
190 |
torch.cuda.empty_cache() # Free up cached memory
|
191 |
+
gc.collect()
|
192 |
time.sleep(0.5)
|
193 |
yield (final_image_path, f"Final image saved at {final_image_path}", iter_images)
|
194 |
else:
|
195 |
torch.cuda.empty_cache() # Free up cached memory
|
196 |
+
gc.collect()
|
197 |
yield (None, "Image generation completed, but no final image was found.", None)
|
198 |
|
199 |
torch.cuda.empty_cache() # Free up cached memory
|
200 |
+
gc.collect()
|
201 |
|
202 |
except torch.cuda.OutOfMemoryError as e:
|
203 |
print(f"Global CUDA Out of Memory Error: {e}")
|