Merge pull request #26 from caetas/dev

ImageNet support and wandb can be disabled

README.md

@@ -161,6 +161,15 @@ The following datasets are ready to be used to train and sample from the provide
 - DTD [`Source`](https://arxiv.org/abs/1311.3618)
 - TinyImageNet [`Source`](https://cs231n.stanford.edu/reports/2015/pdfs/yle_project.pdf) **MANUAL DOWNLOAD REQUIRED** [`Link`](https://www.kaggle.com/datasets/nikhilshingadiya/tinyimagenet200)
 - Horse2Zebra [`Source`](https://arxiv.org/abs/1703.10593v6) **MANUAL DOWNLOAD REQUIRED** [`Link`](https://www.kaggle.com/datasets/balraj98/horse2zebra-dataset)
+- ImageNet-1k [`Source`](https://ieeexplore.ieee.org/abstract/document/5206848) **MANUAL DOWNLOAD REQUIRED** [`Link`](https://www.kaggle.com/c/imagenet-object-localization-challenge/data) **AND PRE-PROCESSING FOLLOWING THE INSTRUCTIONS BELOW**
+
+#### ImageNet-1K Pre-Processing
+
+ImageNet-1K must be downloaded from [Kaggle](https://www.kaggle.com/c/imagenet-object-localization-challenge/data). After downloading it:
+1. Unzip the Folder.
+2. Rename the folder `ILSVRC/Data/CLS-LOC` as `imagenet`.
+3. Move `imagenet` only to [`data/raw`](data/raw). The other folders can be deleted.
+4. Run the [`process_imagenet.py`](src/generativezoo/process_imagenet.py) script. This operation takes a while due to de dataset size.
 
 ## Tracking
 
@@ -178,6 +187,14 @@ The code examples are setup to use [Weights & Biases](https://wandb.ai/home) as
     wandb login
     ```
 
+**If you want to fully disable Weights & Biases during training, use the flag `--no_wandb`.**
+
+**If you want to turn off syncing with the server but want to retain the local copy, run:**
+
+    ```bash
+    wandb offline
+    ```
+
 ## Repository Information
 
 ### Documentation

VERSION

@@ -1 +1 @@
-0.20.0
+1.1.0