Update fault_tolerant_training_basic.rst (#16012)

Author: edenlightning

docs/source-pytorch/clouds/fault_tolerant_training_basic.rst

@@ -18,26 +18,3 @@ With Fault Tolerant Training, when ``Trainer.fit()`` fails in the middle of an e
 Lightning will restart exactly where it failed, and everything will be restored (down to the batch it was on even if the dataset was shuffled).
 
 .. warning:: Fault-tolerant Training is currently an experimental feature within Lightning.
-
-----
-
-***************************************************
-Use fault-tolerance to save money on cloud training
-***************************************************
-Cloud providers offer pre-emptible machines which can be priced as low as 1/10th the cost but can be shut-down automatically at any time.
-Because fault-tolerant training can automatically recover from an interruption, you can train models for many weeks/months at a time for the pre-emptible prices.
-
-To easily run on the cloud with fault-tolerance with lightning-grid, use the following arguments:
-
-.. code-block:: bash
-
-    grid run --use_spot --auto_resume lightning_script.py
-
-The ``--use_spot`` argument enables cheap preemptible pricing (but the machines that can be interrupted).
-If the machine is interrupted, the ``--auto_resume`` argument automatically restarts the machine.
-
-As long as you are running a script that runs a lightning model, the model will restore itself and handle all the details of fault tolerance.
-
-----
-
-.. include:: grid_costs.rst

docs/source-pytorch/clouds/fault_tolerant_training_expert.rst

@@ -21,14 +21,3 @@ To enable fault tolerance on your own cloud or cluster environment enable the *P
 Although Lighting will now be fault-tolerant, you'll have to handle all the nuances of making sure the models are automatically restarted.
 
 .. note:: This complexity is already handled for you if you use **lightning-grid**.
-
-----
-
-**************************************************
-Enable fault-tolerant behavior on your own cluster
-**************************************************
-The simplest way to enable fault-tolerant behavior is to enable lightning-grid to work on your on-prem cluster or cloud environment which will handle all the nuances of fault-tolerant training at scale.
-
-Email us to connect with your own cloud account:
-
-`<[email protected]>`_