Initial commit

.gitattributes

@@ -0,0 +1,28 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zstandard filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+diffusion_model.pt filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,45 @@
+---
+tags:
+- ddpm_diffusion
+---
+
+# Denoising Diffusion Probabilistic Models (DDPM)
+
+**Paper**: [Denoising Diffusion Probabilistic Models](https://arxiv.org/abs/2006.11239)
+
+**Abstract**:
+
+*We present high quality image synthesis results using diffusion probabilistic models, a class of latent variable models inspired by considerations from nonequilibrium thermodynamics. Our best results are obtained by training on a weighted variational bound designed according to a novel connection between diffusion probabilistic models and denoising score matching with Langevin dynamics, and our models naturally admit a progressive lossy decompression scheme that can be interpreted as a generalization of autoregressive decoding. On the unconditional CIFAR10 dataset, we obtain an Inception score of 9.46 and a state-of-the-art FID score of 3.17. On 256x256 LSUN, we obtain sample quality similar to ProgressiveGAN.*
+
+## Usage
+
+```python
+# !pip install diffusers
+from diffusers import DiffusionPipeline
+import PIL.Image
+import numpy as np
+
+model_id = "fusing/ddpm-lsun-bedroom-ema"
+
+# load model and scheduler
+ddpm = DiffusionPipeline.from_pretrained(model_id)
+
+# run pipeline in inference (sample random noise and denoise)
+image = ddpm()
+
+# process image to PIL
+image_processed = image.cpu().permute(0, 2, 3, 1)
+image_processed = (image_processed + 1.0) * 127.5
+image_processed = image_processed.numpy().astype(np.uint8)
+image_pil = PIL.Image.fromarray(image_processed[0])
+
+# save image
+image_pil.save("test.png")
+```
+
+## Samples
+
+1. ![sample_1](https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/hf/ddpm-lsun-bedroom-ema/image_0.png)
+2. ![sample_1](https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/hf/ddpm-lsun-bedroom-ema/image_1.png)
+3. ![sample_1](https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/hf/ddpm-lsun-bedroom-ema/image_2.png)
+4. ![sample_1](https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/hf/ddpm-lsun-bedroom-ema/image_3.png)

config.json

@@ -0,0 +1 @@
+{"_class_name": "UNetModel", "attn_resolutions": [16], "down_blocks": ["UNetResDownBlock2D", "UNetResDownBlock2D", "UNetResDownBlock2D", "UNetResDownBlock2D", "UNetResAttnDownBlock2D", "UNetResDownBlock2D"], "up_blocks": ["UNetResUpBlock2D", "UNetResAttnUpBlock2D", "UNetResUpBlock2D", "UNetResUpBlock2D", "UNetResUpBlock2D", "UNetResUpBlock2D"], "conv_resample": true, "downsample_padding": 0, "num_head_channels": null, "ch": 128, "ch_mult": [1, 1, 2, 2, 4, 4], "block_channels": [128, 128, 256, 256, 512, 512], "resnet_eps": 1e-06, "flip_sin_to_cos": false, "downscale_freq_shift": 1, "dropout": 0.0, "in_channels": 3, "name_or_path": "./ddpm-lsun-church/", "num_res_blocks": 2, "out_ch": 3, "out_channels": 3, "resamp_with_conv": true, "resolution": 256, "image_size": 256}

diffusion_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1296e7d066e7c3c973dc3f3a27c6adca3a981646ede449f127db566da9abcd5
+size 470185705

model_index.json

@@ -0,0 +1,12 @@
+{
+    "_class_name": "DDPM",
+    "_module": "modeling_ddpm.py",
+    "scheduler": [
+      "diffusers",
+      "DDPMScheduler"
+    ],
+    "unet": [
+      "diffusers",
+      "UNetModel"
+    ]
+  }

modeling_ddpm.py

@@ -0,0 +1,61 @@
+# Copyright 2022 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+
+# limitations under the License.
+
+
+from diffusers import DiffusionPipeline
+import tqdm
+import torch
+
+
+class DDPM(DiffusionPipeline):
+
+    modeling_file = "modeling_ddpm.py"
+
+    def __init__(self, unet, noise_scheduler):
+        super().__init__()
+        self.register_modules(unet=unet, noise_scheduler=noise_scheduler)
+
+    def __call__(self, batch_size=1, generator=None, torch_device=None):
+        if torch_device is None:
+            torch_device = "cuda" if torch.cuda.is_available() else "cpu"
+
+        self.unet.to(torch_device)
+        # 1. Sample gaussian noise
+        image = self.noise_scheduler.sample_noise((batch_size, self.unet.in_channels, self.unet.resolution, self.unet.resolution), device=torch_device, generator=generator)
+        for t in tqdm.tqdm(reversed(range(len(self.noise_scheduler))), total=len(self.noise_scheduler)):
+            # i) define coefficients for time step t
+            clip_image_coeff = 1 / torch.sqrt(self.noise_scheduler.get_alpha_prod(t))
+            clip_noise_coeff = torch.sqrt(1 / self.noise_scheduler.get_alpha_prod(t) - 1)
+            image_coeff = (1 - self.noise_scheduler.get_alpha_prod(t - 1)) * torch.sqrt(self.noise_scheduler.get_alpha(t)) / (1 - self.noise_scheduler.get_alpha_prod(t))
+            clip_coeff = torch.sqrt(self.noise_scheduler.get_alpha_prod(t - 1)) * self.noise_scheduler.get_beta(t) / (1 - self.noise_scheduler.get_alpha_prod(t))
+
+            # ii) predict noise residual
+            with torch.no_grad():
+                noise_residual = self.unet(image, t)
+
+            # iii) compute predicted image from residual
+            # See 2nd formula at https://github.com/hojonathanho/diffusion/issues/5#issue-896554416 for comparison
+            pred_mean = clip_image_coeff * image - clip_noise_coeff * noise_residual
+            pred_mean = torch.clamp(pred_mean, -1, 1)
+            prev_image = clip_coeff * pred_mean + image_coeff * image
+
+            # iv) sample variance
+            prev_variance = self.noise_scheduler.sample_variance(t, prev_image.shape, device=torch_device, generator=generator)
+
+            # v) sample  x_{t-1} ~ N(prev_image, prev_variance)
+            sampled_prev_image = prev_image + prev_variance
+            image = sampled_prev_image
+
+        return image

scheduler_config.json

@@ -0,0 +1,8 @@
+{
+  "_class_name": "GaussianDDPMScheduler",
+  "beta_end": 0.02,
+  "beta_schedule": "linear",
+  "beta_start": 0.0001,
+  "timesteps": 1000,
+  "variance_type": "fixed_small"
+}