[docs] Move all /latest links to /master (#11897)

* use master link

* remae

* revert non-ray

* more

* mre

python/asv.conf.json

@@ -7,7 +7,7 @@
     "project": "ray",
 
     // The project's homepage
-    "project_url": "http://docs.ray.io/en/latest/index.html",
+    "project_url": "http://docs.ray.io/en/master/index.html",
 
     // The URL or local path of the source code repository for the
     // project being benchmarked

python/ray/dashboard/client/src/pages/dashboard/logical-view/Actor.tsx

@@ -14,7 +14,7 @@ import { sum } from "../../../common/util";
 import ActorDetailsPane from "./ActorDetailsPane";
 
 const memoryDebuggingDocLink =
-  "https://docs.ray.io/en/latest/memory-management.html#debugging-using-ray-memory";
+  "https://docs.ray.io/en/master/memory-management.html#debugging-using-ray-memory";
 
 const useActorStyles = makeStyles((theme: Theme) =>
   createStyles({

python/ray/dashboard/client/src/pages/dashboard/tune/Tune.tsx

@@ -143,7 +143,7 @@ class Tune extends React.Component<
             You can use this tab to monitor Tune jobs, their statuses,
             hyperparameters, and more. For more information, read the
             documentation{" "}
-            <a href="https://docs.ray.io/en/latest/ray-dashboard.html#tune">
+            <a href="https://docs.ray.io/en/master/ray-dashboard.html#tune">
               here
             </a>
             .

python/ray/tune/README.rst

@@ -3,7 +3,7 @@ Tune: Scalable Hyperparameter Tuning
 
 Tune is a scalable framework for hyperparameter search with a focus on deep learning and deep reinforcement learning.
 
-User documentation can be `found here <http://docs.ray.io/en/latest/tune.html>`__.
+User documentation can be `found here <http://docs.ray.io/en/master/tune.html>`__.
 
 
 Tutorial

python/ray/tune/tests/ext_pytorch.py

@@ -17,7 +17,7 @@ accurate one. Often simple things like choosing a different learning rate or cha
 a network layer size can have a dramatic impact on your model performance.
 
 Fortunately, there are tools that help with finding the best combination of parameters.
-`Ray Tune <https://docs.ray.io/en/latest/tune.html>`_ is an industry standard tool for
+`Ray Tune <https://docs.ray.io/en/master/tune.html>`_ is an industry standard tool for
 distributed hyperparameter tuning. Ray Tune includes the latest hyperparameter search
 algorithms, integrates with TensorBoard and other analysis libraries, and natively
 supports distributed training through `Ray's distributed machine learning engine

python/ray/util/joblib/__init__.py

@@ -9,7 +9,7 @@ def register_ray():
     except ImportError:
         msg = ("To use the ray backend you must install ray."
                "Try running 'pip install ray'."
-               "See https://docs.ray.io/en/latest/installation.html"
+               "See https://docs.ray.io/en/master/installation.html"
                "for more information.")
         raise ImportError(msg)
 

python/ray/util/sgd/torch/examples/benchmarks/README.rst

@@ -3,7 +3,7 @@ Running benchmarks
 
 RaySGD provides comparable or better performance than other existing solutions for parallel or distributed training.
 
-You can run ``ray/python/ray/util/sgd/torch/examples/benchmarks/benchmark.py`` for benchmarking the RaySGD TorchTrainer implementation. To benchmark training on a multi-node multi-gpu cluster, you can use the `Ray Autoscaler <https://docs.ray.io/en/latest/autoscaling.html#aws>`_.
+You can run ``ray/python/ray/util/sgd/torch/examples/benchmarks/benchmark.py`` for benchmarking the RaySGD TorchTrainer implementation. To benchmark training on a multi-node multi-gpu cluster, you can use the `Ray Autoscaler <https://docs.ray.io/en/master/autoscaling.html#aws>`_.
 
 DISCLAIMER: RaySGD does not provide any custom communication primitives. If you see any performance issues, you may need to file them on the PyTorch github repository.