diff --git a/.flake8 b/.flake8
index 0cc61b77a7..9ce25d4544 100644
--- a/.flake8
+++ b/.flake8
@@ -6,4 +6,6 @@ ignore = W503, E203, E221, C901, C408, E741
 max-line-length = 100
 max-complexity = 18
 select = B,C,E,F,W,T4,B9
-exclude = build,__init__.py
+exclude = build
+per-file-ignores =
+	**/__init__.py:F401,F403
diff --git a/INSTALL.md b/INSTALL.md
index 3985f8ae4f..a4b4d406e9 100644
--- a/INSTALL.md
+++ b/INSTALL.md
@@ -35,28 +35,27 @@ old build first. You often need to rebuild detectron2 after reinstalling PyTorch
 
 ### Install Pre-Built Detectron2 (Linux only)
 ```
-# for CUDA 10.1:
-python -m pip install detectron2 -f https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/index.html
+# for CUDA 10.2:
+python -m pip install detectron2 -f https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/index.html
 ```
-You can replace cu101 with "cu{100,92}" or "cpu".
+For other cuda versions, replace cu102 with "cu{101,92}" or "cpu".
 
 Note that:
 1. Such installation has to be used with certain version of official PyTorch release.
    See [releases](https://github.com/facebookresearch/detectron2/releases) for requirements.
    It will not work with a different version of PyTorch or a non-official build of PyTorch.
+	 The CUDA version used by PyTorch and detectron2 has to match as well.
 2. Such installation is out-of-date w.r.t. master branch of detectron2. It may not be
 	 compatible with the master branch of a research project that uses detectron2 (e.g. those in
 	 [projects](projects) or [meshrcnn](https://github.com/facebookresearch/meshrcnn/)).
 
 ### Common Installation Issues
 
-If you met issues using the pre-built detectron2, please uninstall it and try building it from source.
-
 Click each issue for its solutions:
 
 <details>
 <summary>
-Undefined torch/aten/caffe2 symbols, or segmentation fault immediately when running the library.
+Undefined torch/aten/caffe2 symbols; missing torch dynamic libraries; segmentation fault immediately when using detectron2.
 </summary>
 <br/>
 
@@ -144,7 +143,7 @@ Two possibilities:
 
 <details>
 <summary>
-Undefined CUDA symbols; cannot open libcudart.so; other nvcc failures.
+Undefined CUDA symbols; cannot open libcudart.so
 </summary>
 <br/>
 The version of NVCC you use to build detectron2 or torchvision does
@@ -161,6 +160,18 @@ to match your local CUDA installation, or install a different version of CUDA to
 </details>
 
 
+<details>
+<summary>
+C++ compilation errors from NVCC
+</summary>
+<br/>
+1. NVCC version has to match the CUDA version of your PyTorch.
+
+2. NVCC has compatibility issues with certain versions of gcc. You may need a different
+	 version of gcc. The version used by PyTorch can be found by `print(torch.__config__.show())`.
+</details>
+
+
 <details>
 <summary>
 "ImportError: cannot import name '_C'".
diff --git a/detectron2/config/defaults.py b/detectron2/config/defaults.py
index 88bb89ed49..76567c813a 100644
--- a/detectron2/config/defaults.py
+++ b/detectron2/config/defaults.py
@@ -28,7 +28,7 @@
 _C.MODEL.DEVICE = "cuda"
 _C.MODEL.META_ARCHITECTURE = "GeneralizedRCNN"
 
-# Path (possibly with schema like catalog:// or detectron2://) to a checkpoint file
+# Path (a file path, or URL like detectron2://.., https://..) to a checkpoint file
 # to be loaded to the model. You can find available models in the model zoo.
 _C.MODEL.WEIGHTS = ""
 
diff --git a/detectron2/data/transforms/transform_gen.py b/detectron2/data/transforms/transform_gen.py
index 197a0ebf67..3c9920db5b 100644
--- a/detectron2/data/transforms/transform_gen.py
+++ b/detectron2/data/transforms/transform_gen.py
@@ -446,7 +446,7 @@ def get_transform(self, img):
 
 class RandomSaturation(TransformGen):
     """
-    Randomly transforms image saturation.
+    Randomly transforms saturation of an RGB image.
 
     Saturation intensity is uniformly sampled in (intensity_min, intensity_max).
     - intensity < 1 will reduce saturation (make the image more grayscale)
@@ -466,7 +466,7 @@ def __init__(self, intensity_min, intensity_max):
         self._init(locals())
 
     def get_transform(self, img):
-        assert img.shape[-1] == 3, "Saturation only works on RGB images"
+        assert img.shape[-1] == 3, "RandomSaturation only works on RGB images"
         w = np.random.uniform(self.intensity_min, self.intensity_max)
         grayscale = img.dot([0.299, 0.587, 0.114])[:, :, np.newaxis]
         return BlendTransform(src_image=grayscale, src_weight=1 - w, dst_weight=w)
@@ -474,7 +474,8 @@ def get_transform(self, img):
 
 class RandomLighting(TransformGen):
     """
-    Randomly transforms image color using fixed PCA over ImageNet.
+    The "lighting" augmentation described in AlexNet, using fixed PCA over ImageNet.
+    Inputs are assumed to be RGB images.
 
     The degree of color jittering is randomly sampled via a normal distribution,
     with standard deviation given by the scale parameter.
@@ -493,7 +494,7 @@ def __init__(self, scale):
         self.eigen_vals = np.array([0.2175, 0.0188, 0.0045])
 
     def get_transform(self, img):
-        assert img.shape[-1] == 3, "Saturation only works on RGB images"
+        assert img.shape[-1] == 3, "RandomLighting only works on RGB images"
         weights = np.random.normal(scale=self.scale, size=3)
         return BlendTransform(
             src_image=self.eigen_vecs.dot(weights * self.eigen_vals), src_weight=1.0, dst_weight=1.0
diff --git a/detectron2/modeling/__init__.py b/detectron2/modeling/__init__.py
index 9e23fe4a70..1fc83c9003 100644
--- a/detectron2/modeling/__init__.py
+++ b/detectron2/modeling/__init__.py
@@ -1,6 +1,4 @@
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
-import torch
-
 from detectron2.layers import ShapeSpec
 
 from .anchor_generator import build_anchor_generator, ANCHOR_GENERATOR_REGISTRY
@@ -48,9 +46,5 @@
 )
 from .test_time_augmentation import DatasetMapperTTA, GeneralizedRCNNWithTTA
 
-_EXCLUDE = {"torch", "ShapeSpec"}
+_EXCLUDE = {"ShapeSpec"}
 __all__ = [k for k in globals().keys() if k not in _EXCLUDE and not k.startswith("_")]
-
-assert (
-    torch.Tensor([1]) == torch.Tensor([2])
-).dtype == torch.bool, "Your Pytorch is too old. Please update to contain https://github.com/pytorch/pytorch/pull/21113"
diff --git a/docs/tutorials/models.md b/docs/tutorials/models.md
index 456f36d1c0..3c635747ad 100644
--- a/docs/tutorials/models.md
+++ b/docs/tutorials/models.md
@@ -14,7 +14,7 @@ and how to use the `model` object.
 ### Load/Save a Checkpoint
 ```python
 from detectron2.checkpoint import DetectionCheckpointer
-DetectionCheckpointer(model).load(file_path)   # load a file to model
+DetectionCheckpointer(model).load(file_path_or_url)  # load a file, usually from cfg.MODEL.WEIGHTS
 
 checkpointer = DetectionCheckpointer(model, save_dir="output")
 checkpointer.save("model_999")  # save to output/model_999.pth
@@ -76,9 +76,9 @@ The dict may contain the following keys:
   + "proposal_boxes": a [Boxes](../modules/structures.html#detectron2.structures.Boxes) object storing P proposal boxes.
   + "objectness_logits": `Tensor`, a vector of P scores, one for each proposal.
 * "height", "width": the **desired** output height and width, which is not necessarily the same
-  as the height or width of the `image` input field.
-  For example, the `image` input field might be a resized image,
-  but you may want the outputs to be in **original** resolution.
+  as the height or width of the `image` field.
+  For example, the `image` field contains the resized image, if resize is used as a preprocessing step.
+  But you may want the outputs to be in **original** resolution.
 
   If provided, the model will produce output in this resolution,
   rather than in the resolution of the `image` as input into the model. This is more efficient and accurate.
diff --git a/docs/tutorials/write-models.md b/docs/tutorials/write-models.md
index bb87d586d6..8ed1f2425b 100644
--- a/docs/tutorials/write-models.md
+++ b/docs/tutorials/write-models.md
@@ -13,6 +13,7 @@ from detectron2.modeling import BACKBONE_REGISTRY, Backbone, ShapeSpec
 @BACKBONE_REGISTRY.register()
 class ToyBackBone(Backbone):
   def __init__(self, cfg, input_shape):
+    super().__init__()
     # create your own backbone
     self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=16, padding=3)
 
diff --git a/projects/PointRend/README.md b/projects/PointRend/README.md
index 443736fff3..1f3659367d 100644
--- a/projects/PointRend/README.md
+++ b/projects/PointRend/README.md
@@ -11,7 +11,7 @@ Alexander Kirillov, Yuxin Wu, Kaiming He, Ross Girshick
 In this repository, we release code for PointRend in Detectron2. PointRend can be flexibly applied to both instance and semantic segmentation tasks by building on top of existing state-of-the-art models.
 
 ## Installation
-Install Detectron 2 following [INSTALL.md](https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md). You are ready to go!
+Install Detectron2 following [the instructions](https://detectron2.readthedocs.io/tutorials/install.html). You are ready to go!
 
 ## Quick start and visualization
 
diff --git a/tools/train_net.py b/tools/train_net.py
index da8c7e91fc..ad568c0df2 100755
--- a/tools/train_net.py
+++ b/tools/train_net.py
@@ -44,9 +44,8 @@ class Trainer(DefaultTrainer):
     """
     We use the "DefaultTrainer" which contains pre-defined default logic for
     standard training workflow. They may not work for you, especially if you
-    are working on a new research project. In that case you can use the cleaner
-    "SimpleTrainer", or write your own training loop. You can use
-    "tools/plain_train_net.py" as an example.
+    are working on a new research project. In that case you can write your
+    own training loop. You can use "tools/plain_train_net.py" as an example.
     """
 
     @classmethod