The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Fix quality loss due to low precision
City9bfec2bdbf0b
diff --git a/comfy/sd.py b/comfy/sd.pyindex 2f1b2e964..f186273ea 100644--- a/comfy/sd.py+++ b/comfy/sd.py@@ -183,7 +183,7 @@ class VAE:steps += pixel_samples.shape[0] * comfy.utils.get_tiled_scale_steps(pixel_samples.shape[3], pixel_samples.shape[2], tile_x * 2, tile_y // 2, overlap)pbar = comfy.utils.ProgressBar(steps)- encode_fn = lambda a: self.first_stage_model.encode(2. * a.to(self.vae_dtype).to(self.device) - 1.).sample().float()+ encode_fn = lambda a: self.first_stage_model.encode((2. * a - 1.).to(self.vae_dtype).to(self.device)).sample().float()samples = comfy.utils.tiled_scale(pixel_samples, encode_fn, tile_x, tile_y, overlap, upscale_amount = (1/8), out_channels=4, pbar=pbar)samples += comfy.utils.tiled_scale(pixel_samples, encode_fn, tile_x * 2, tile_y // 2, overlap, upscale_amount = (1/8), out_channels=4, pbar=pbar)samples += comfy.utils.tiled_scale(pixel_samples, encode_fn, tile_x // 2, tile_y * 2, overlap, upscale_amount = (1/8), out_channels=4, pbar=pbar)@@ -202,7 +202,7 @@ class VAE:pixel_samples = torch.empty((samples_in.shape[0], 3, round(samples_in.shape[2] * 8), round(samples_in.shape[3] * 8)), device="cpu")for x in range(0, samples_in.shape[0], batch_number):samples = samples_in[x:x+batch_number].to(self.vae_dtype).to(self.device)- pixel_samples[x:x+batch_number] = torch.clamp((self.first_stage_model.decode(samples) + 1.0) / 2.0, min=0.0, max=1.0).cpu().float()+ pixel_samples[x:x+batch_number] = torch.clamp((self.first_stage_model.decode(samples).cpu().float() + 1.0) / 2.0, min=0.0, max=1.0)except model_management.OOM_EXCEPTION as e:print("Warning: Ran out of memory when regular VAE decoding, retrying with tiled VAE decoding.")pixel_samples = self.decode_tiled_(samples_in)