SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

doc: update W&B related doc (#10086)

Browse Source 
* doc: update W&B related doc

* doc(wandb): mention report_to

* doc(wandb): commit suggestion

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* doc(wandb): fix typo

* doc(wandb): remove WANDB_DISABLED

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
This commit is contained in:
Boris Dayma
2021-02-09 13:47:52 -06:00
committed by GitHub
parent 480a9d6ba0
commit 7c7962ba89
3 changed files with 42 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
examples/README.md
View File

@@ -1,6 +1,5 @@
<!---
Copyright 2020 The HuggingFace Team. All rights reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
@@ -147,7 +146,7 @@ python xla_spawn.py --num_cores 8 \
You can easily log and monitor your runs code. The following are currently supported:
* [TensorBoard](https://www.tensorflow.org/tensorboard)
* [Weights & Biases](https://docs.wandb.com/library/integrations/huggingface)
* [Weights & Biases](https://docs.wandb.ai/integrations/huggingface)
* [Comet ML](https://www.comet.ml/docs/python-sdk/huggingface/)
### Weights & Biases
@@ -171,9 +170,46 @@ import wandb
wandb.login()
```
To enable logging to W&B, include `"wandb"` in the `report_to` of your `TrainingArguments` or script. Or just pass along `--report_to all` if you have `wandb` installed.
Whenever you use `Trainer` or `TFTrainer` classes, your losses, evaluation metrics, model topology and gradients (for `Trainer` only) will automatically be logged.
When using 🤗 Transformers with PyTorch Lightning, runs can be tracked through `WandbLogger`. Refer to related [documentation & examples](https://docs.wandb.com/library/integrations/lightning).
Advanced configuration is possible by setting environment variables:
<table>
<thead>
<tr>
<th style="text-align:left">Environment Variables</th>
<th style="text-align:left">Options</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align:left">WANDB_LOG_MODEL</td>
<td style="text-align:left">Log the model as artifact at the end of training (<b>false</b> by default)</td>
</tr>
<tr>
<td style="text-align:left">WANDB_WATCH</td>
<td style="text-align:left">
<ul>
<li><b>gradients</b> (default): Log histograms of the gradients</li>
<li><b>all</b>: Log histograms of gradients and parameters</li>
<li><b>false</b>: No gradient or parameter logging</li>
</ul>
</td>
</tr>
<tr>
<td style="text-align:left">WANDB_PROJECT</td>
<td style="text-align:left">Organize runs by project</td>
</tr>
</tbody>
</table>
Set run names with `run_name` argument present in scripts or as part of `TrainingArguments`.
Additional configuration options are available through generic [wandb environment variables](https://docs.wandb.com/library/environment-variables).
Refer to related [documentation & examples](https://docs.wandb.ai/integrations/huggingface).
### Comet.ml