Update copyright year on demo Diffusion

Signed-off-by: Shuyue Lan <[email protected]>

CHANGELOG.md

@@ -1,6 +1,6 @@
 # TensorRT OSS Release Changelog
 
-## [8.6.0 EA](https://docs.nvidia.com/deeplearning/tensorrt/release-notes/tensorrt-8.html#tensorrt-8) - 2023-03-10
+## [8.6.0 EA](https://docs.nvidia.com/deeplearning/tensorrt/release-notes/tensorrt-8.html#tensorrt-8) - 2023-03-14
 
 TensorRT OSS release corresponding to TensorRT 8.6.0.12 EA release.
 - Updates since [TensorRT 8.5.3 GA release](https://docs.nvidia.com/deeplearning/tensorrt/release-notes/tensorrt-8.html#rel-8-5-3).

demo/Diffusion/demo_img2img.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");

demo/Diffusion/demo_inpaint.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -55,13 +55,13 @@ def parseArgs():
     else:
         img_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png"
         input_image = download_image(img_url)
-    
+
     if args.mask_image:
         mask_image = Image.open(args.mask_image).convert("RGB")
     else:
         mask_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png"
         mask_image = download_image(mask_url)
-    
+
     image_width, image_height = input_image.size
     mask_width, mask_height = mask_image.size
 
@@ -96,7 +96,7 @@ def parseArgs():
         max_batch_size=max_batch_size)
 
     # Load TensorRT engines and pytorch modules
-    demo.loadEngines(args.engine_dir, args.onnx_dir, args.onnx_opset, 
+    demo.loadEngines(args.engine_dir, args.onnx_dir, args.onnx_opset,
         opt_batch_size=len(prompt), opt_image_height=image_height, opt_image_width=image_width, \
         force_export=args.force_onnx_export, force_optimize=args.force_onnx_optimize, \
         force_build=args.force_engine_build, \

demo/Diffusion/demo_txt2img.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -73,7 +73,7 @@ def parseArgs():
         max_batch_size=max_batch_size)
 
     # Load TensorRT engines and pytorch modules
-    demo.loadEngines(args.engine_dir, args.onnx_dir, args.onnx_opset, 
+    demo.loadEngines(args.engine_dir, args.onnx_dir, args.onnx_opset,
         opt_batch_size=len(prompt), opt_image_height=image_height, opt_image_width=image_width, \
         force_export=args.force_onnx_export, force_optimize=args.force_onnx_optimize, \
         force_build=args.force_engine_build, \

demo/Diffusion/img2img_pipeline.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");

demo/Diffusion/inpaint_pipeline.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -42,7 +42,7 @@ def __init__(
 
         if scheduler != "PNDM":
             raise ValueError(f"Inpainting only supports PNDM scheduler")
-        
+
         super(InpaintPipeline, self).__init__(*args, **kwargs, \
             inpaint=True, scheduler=scheduler, stages=[ 'vae_encoder', 'clip', 'unet', 'vae'])
 
@@ -132,4 +132,3 @@ def infer(
             if not warmup:
                 self.print_summary(self.denoising_steps, e2e_tic, e2e_toc, vae_enc=True)
                 self.save_image(images, 'inpaint', prompt)
-

demo/Diffusion/stable_diffusion_pipeline.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");

demo/Diffusion/txt2img_pipeline.py

@@ -1,5 +1,5 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -72,7 +72,7 @@ def infer(
                 Verbose in logging
         """
         assert len(prompt) == len(negative_prompt)
-    
+
         with torch.inference_mode(), torch.autocast("cuda"), trt.Runtime(TRT_LOGGER):
             # Pre-initialize latents
             latents = self.initialize_latents( \

demo/Diffusion/utilities.py

@@ -1,6 +1,6 @@
 #
 # Copyright 2022 The HuggingFace Inc. team.
-# SPDX-FileCopyrightText: Copyright (c) 1993-2022 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -711,7 +711,7 @@ def configure(self):
 
             timestep_list = [self.timesteps[step_index - 1], timestep]
             self.multistep_dpm_solver_second_order_coefs_precompute(timestep_list, prev_timestep)
-                
+
             timestep_list = [self.timesteps[step_index - 2], self.timesteps[step_index - 1], timestep]
             self.multistep_dpm_solver_third_order_coefs_precompute(timestep_list, prev_timestep)
 
@@ -738,7 +738,7 @@ def multistep_dpm_solver_second_order_coefs_precompute(self, timestep_list, prev
         h = lambda_t - lambda_s0
         if self.algorithm_type == "dpmsolver++":
             # See https://arxiv.org/abs/2211.01095 for detailed derivations
-            if self.solver_type == "midpoint": 
+            if self.solver_type == "midpoint":
                 self.second_order_first_coef.append(sigma_t / sigma_s0)
                 self.second_order_second_coef.append((alpha_t * (torch.exp(-h) - 1.0)))
                 self.second_order_third_coef.append(0.5 * (alpha_t * (torch.exp(-h) - 1.0)))
@@ -1011,7 +1011,7 @@ def __init__(
         alphas = 1.0 - betas
         self.alphas_cumprod = torch.cumprod(alphas, dim=0).to(device=self.device)
         self.final_alpha_cumprod = self.alphas_cumprod[0]
-        
+
         # standard deviation of the initial noise distribution
         self.init_noise_sigma = 1.0
         self.steps_offset = steps_offset
@@ -1041,7 +1041,7 @@ def set_timesteps(self, steps):
 
     def scale_model_input(self, sample: torch.FloatTensor, idx, *args, **kwargs) -> torch.FloatTensor:
         return sample
-    
+
     def configure(self):
         self.alphas_cumprod_prev = torch.roll(self.alphas_cumprod, shifts=self.step_ratio)
         self.alphas_cumprod_prev[:self.step_ratio] = self.final_alpha_cumprod
@@ -1063,7 +1063,7 @@ def configure(self):
     def step(self, output, sample, idx, timestep):
         # step_plms: propagate the sample with the linear multi-step method. This has one forward pass with multiple
         # times to approximate the solution.
-        
+
         # prev_timestep = timestep - self.step_ratio
 
         if self.counter != 1:
@@ -1086,7 +1086,7 @@ def step(self, output, sample, idx, timestep):
             output = (23 * self.ets[-1] - 16 * self.ets[-2] + 5 * self.ets[-3]) / 12
         else:
             output = (1 / 24) * (55 * self.ets[-1] - 59 * self.ets[-2] + 37 * self.ets[-3] - 9 * self.ets[-4])
-        
+
         if self.prediction_type == "v_prediction":
             output = (self.alphas_cumprod[idx]**0.5) * output + (self.beta_cumprod[idx]**0.5) * sample
         elif self.prediction_type != "epsilon":
@@ -1191,5 +1191,3 @@ def add_arguments(parser):
     parser.add_argument('--hf-token', type=str, help="HuggingFace API access token for downloading model checkpoints")
     parser.add_argument('-v', '--verbose', action='store_true', help="Show verbose output")
     return parser
-
-