Upload README.md with huggingface_hub

README.md

@@ -1,54 +1,59 @@
 ---
-language: en
 license: apache-2.0
-library_name: diffusers
-tags: []
-datasets: huggan/flowers-102-categories
-metrics: []
+tags:
+- pytorch
+- diffusers
+- unconditional-image-generation
 ---
 
-<!-- This model card has been generated automatically according to the information the training script had access to. You
-should probably proofread and complete it, then remove this comment. -->
+# Denoising Diffusion Probabilistic Models (DDPM)
 
-# ddpm-ema-flowers-64
+**Paper**: [Denoising Diffusion Probabilistic Models](https://arxiv.org/abs/2006.11239)
 
-## Model description
+**Authors**: Jonathan Ho, Ajay Jain, Pieter Abbeel
 
-This diffusion model is trained with the [🤗 Diffusers](https://github.com/huggingface/diffusers) library 
-on the `huggan/flowers-102-categories` dataset.
+**Abstract**:
 
-## Intended uses & limitations
+*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.*
 
-#### How to use
+## Inference
+
+**DDPM** models can use *discrete noise schedulers* such as:
+
+- [scheduling_ddpm](https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_ddpm.py)
+- [scheduling_ddim](https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_ddim.py)
+- [scheduling_pndm](https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_pndm.py)
+
+for inference. Note that while the *ddpm* scheduler yields the highest quality, it also takes the longest.
+For a good trade-off between quality and inference speed you might want to consider the *ddim* or *pndm* schedulers instead.
+
+See the following code:
 
 ```python
-# TODO: add an example code snippet for running this diffusion pipeline
-```
+# !pip install diffusers
+from diffusers import DDPMPipeline, DDIMPipeline, PNDMPipeline
 
-#### Limitations and bias
+model_id = "google/ddpm-cifar10-32"
 
-[TODO: provide examples of latent issues and potential remediations]
+# load model and scheduler
+ddpm = DDPMPipeline.from_pretrained(model_id)  # you can replace DDPMPipeline with DDIMPipeline or PNDMPipeline for faster inference
 
-## Training data
+# run pipeline in inference (sample random noise and denoise)
+image = ddpm()["sample"]
 
-[TODO: describe the data used to train the model]
 
-### Training hyperparameters
+# save image
+image[0].save("ddpm_generated_image.png")
+```
 
-The following hyperparameters were used during training:
-- learning_rate: 0.0001
-- train_batch_size: 16
-- eval_batch_size: 16
-- gradient_accumulation_steps: 1
-- optimizer: AdamW with betas=(0.95, 0.999), weight_decay=1e-06 and epsilon=1e-08
-- lr_scheduler: cosine
-- lr_warmup_steps: 500
-- ema_inv_gamma: 1.0
-- ema_inv_gamma: 0.75
-- ema_inv_gamma: 0.9999
-- mixed_precision: no
+For more in-detail information, please have a look at the [official inference example](_) # <- TODO(PVP) add link
 
-### Training results
+## Training
 
-📈 [TensorBoard logs](https://huggingface.co/anton-l/ddpm-ema-flowers-64/tensorboard?#scalars)
+If you want to train your own model, please have a look at the [official training example]( ) # <- TODO(PVP) add link
 
+## Samples
+1. ![sample_1](https://huggingface.co/anton-l/ddpm-ema-flowers-64/resolve/main/generated_image_0.png)
+2. ![sample_2](https://huggingface.co/anton-l/ddpm-ema-flowers-64/resolve/main/generated_image_1.png)
+3. ![sample_3](https://huggingface.co/anton-l/ddpm-ema-flowers-64/resolve/main/generated_image_2.png)
+4. ![sample_4](https://huggingface.co/anton-l/ddpm-ema-flowers-64/resolve/main/generated_image_3.png)