From 917dd7e570940d23285e8c08e06b27962e8e2b46 Mon Sep 17 00:00:00 2001
From: Katia Patkin <87335417+kptkin@users.noreply.github.com>
Date: Thu, 11 Aug 2022 15:20:51 -0700
Subject: [PATCH] remove service user doc (#4088)

---
 docs/dev/wandb-service-user.md | 91 ----------------------------------
 1 file changed, 91 deletions(-)
 delete mode 100644 docs/dev/wandb-service-user.md

diff --git a/docs/dev/wandb-service-user.md b/docs/dev/wandb-service-user.md
deleted file mode 100644
index 3155e00f735..00000000000
--- a/docs/dev/wandb-service-user.md
+++ /dev/null
@@ -1,91 +0,0 @@
-<p align="center">
-  <img src="../../.github/wb-logo-lightbg.png#gh-light-mode-only" width="600" alt="Weights & Biases"/>
-  <img src="../../.github/wb-logo-darkbg.png#gh-dark-mode-only" width="600" alt="Weights & Biases"/>
-</p>
-
-# `wandb` service
-
-We are excited :tada: to share with you a new feature we call `service`. This feature is still in the experimental stage, but will be integrated by default in `wandb` in the near future.
-
-## Why would you use this feature?
-
-`service` improves `wandb`'s handling of multiprocessing and thus improves reliability in a distributed training setting.
-If you are using `wandb` in a distributed training setup and experiencing hangs, please consider trying out this new feature.
-
-## Usage
-
-### General usage
-
-`service` can be enabled by adding the following to your script:
-
-```python
-import wandb
-
-if __name__ == "__main__":
-    wandb.require(experiment="service")
-    # <rest-of-your-script-goes-here>
-```
-
-### Advanced usage example
-
-If you are calling `wandb.init` in a spawned process you should add `wandb.setup()` in the main process:
-
-```python
-import multiprocessing as mp
-import wandb
-
-
-def do_work(n):
-    run = wandb.init(config=dict(n=n))
-    run.log(dict(this=n * n))
-
-
-def main():
-    wandb.require("service")
-    wandb.setup()
-    pool = mp.Pool(processes=4)
-    pool.map(do_work, range(4))
-
-
-if __name__ == "__main__":
-    main()
-```
-
-### PyTorch Lightning
-
-If you are using `PyTorch Lightning` please:
-
-- Install the most recent version of `wandb` and update `PyTorch Lightning` from the master branch:
-
-```bash
-pip install --upgrade wandb
-pip install --upgrade git+https://github.com/PytorchLightning/pytorch-lightning.git
-```
-
-The feature support will be part of `PyTorch Lightning`'s [1.6.0 release](https://github.com/PyTorchLightning/pytorch-lightning/pull/11650), at which point you will be able to simply install/update it from PyPI.
-
-That's it... no need to add any further code...
-
-## FAQs
-
-### If your script is stuck in a restart loop
-
-Please try adding:
-
-```python
-if __name__ == "__main__":
-    # <your-script-goes-here>
-    ...
-```
-
-## Reporting issues
-
-We appreciate it that you gave this feature a try. If you are experiencing an issue that is not listed in the FAQs, please file a [GitHub Issue](https://github.com/wandb/wandb/issues).
-To help us reproduce your issue, please provide the following:
-
-- Minimal script to reproduce
-- Python version
-- Operating System
-- Platform (for example: GPUs, TPUs etc.)
-- Additional relevant packages installed (you can for example run `pip list` in your python environment)
-- Traceback (if relevant)