massive img downscaling for fast sending...

This commit is contained in:
Johannes Stelzer 2023-01-12 10:48:04 +01:00
parent 30f4aaaa24
commit b0555f1954
1 changed files with 14 additions and 6 deletions

View File

@ -67,14 +67,15 @@ class BlendingFrontend():
self.nmb_imgs_show = 5 self.nmb_imgs_show = 5
self.fps = 30 self.fps = 30
self.duration = 10 self.duration = 10
self.max_size_imgs = 200 # gradio otherwise mega slow
if not self.use_debug: if not self.use_debug:
self.lb.sdh.num_inference_steps = self.num_inference_steps self.lb.sdh.num_inference_steps = self.num_inference_steps
self.height = self.lb.sdh.height self.height = self.lb.sdh.height
self.width = self.lb.sdh.width self.width = self.lb.sdh.width
else: else:
self.height = 420 self.height = 768
self.width = 420 self.width = 768
def change_depth_strength(self, value): def change_depth_strength(self, value):
self.depth_strength = value self.depth_strength = value
@ -150,6 +151,10 @@ class BlendingFrontend():
print(f"randomize_seed2: new seed = {self.seed2}") print(f"randomize_seed2: new seed = {self.seed2}")
return seed return seed
def downscale_imgs(self, list_imgs):
return [l.resize((self.max_size_imgs, self.max_size_imgs)) for l in list_imgs]
def run(self, x): def run(self, x):
print("STARTING DIFFUSION!") print("STARTING DIFFUSION!")
self.state_prev = self.state_current.copy() self.state_prev = self.state_current.copy()
@ -159,7 +164,10 @@ class BlendingFrontend():
if self.use_debug: if self.use_debug:
list_imgs = [(255*np.random.rand(self.height,self.width,3)).astype(np.uint8) for l in range(5)] list_imgs = [(255*np.random.rand(self.height,self.width,3)).astype(np.uint8) for l in range(5)]
list_imgs = [Image.fromarray(l) for l in list_imgs]
list_imgs = self.downscale_imgs(list_imgs)
self.imgs_show_current = copy.deepcopy(list_imgs) self.imgs_show_current = copy.deepcopy(list_imgs)
print("DONE! SENDING BACK RESULTS")
return list_imgs return list_imgs
self.lb.set_width(self.width) self.lb.set_width(self.width)
@ -190,8 +198,10 @@ class BlendingFrontend():
list_imgs = [] list_imgs = []
for j in idx_list: for j in idx_list:
list_imgs.append(imgs_transition[j]) list_imgs.append(imgs_transition[j])
self.imgs_show_current = copy.deepcopy(list_imgs)
list_imgs = self.downscale_imgs(list_imgs)
self.imgs_show_current = copy.deepcopy(list_imgs)
return list_imgs return list_imgs
@ -261,9 +271,7 @@ class BlendingFrontend():
if __name__ == "__main__": if __name__ == "__main__":
fp_ckpt = "../stable_diffusion_models/ckpt/v2-1_512-ema-pruned.ckpt" fp_ckpt = "../stable_diffusion_models/ckpt/v2-1_512-ema-pruned.ckpt"
fp_config = 'configs/v2-inference.yaml' sdh = StableDiffusionHolder(fp_ckpt)
sdh = StableDiffusionHolder(fp_ckpt, fp_config)
self = BlendingFrontend(sdh) self = BlendingFrontend(sdh)