Skip to content

Commit

Permalink
Update deepspeed.md (#2286)
Browse files Browse the repository at this point in the history
  • Loading branch information
stas00 authored Dec 27, 2023
1 parent 3db088f commit ad957ce
Showing 1 changed file with 7 additions and 0 deletions.
7 changes: 7 additions & 0 deletions docs/source/usage_guides/deepspeed.md
Original file line number Diff line number Diff line change
Expand Up @@ -422,6 +422,13 @@ We will look at the changes needed in the code when using these.
based on model, dataloaders, dummy optimizer and dummy schedulers provided to `prepare` method.
Only the `auto` fields specified in above examples are handled by `prepare` method and the rest have to be explicitly specified by the user.

The `auto` values are calculated as:

- `reduce_bucket_size`: `hidden_size*hidden_size`
- `stage3_prefetch_bucket_size`: `0.9 * hidden_size * hidden_size`
- `stage3_param_persistence_threshold`: `10 * hidden_size`


**Things to note when using DeepSpeed Config File**

Below is a sample script using `deepspeed_config_file` in different scenarios.
Expand Down

0 comments on commit ad957ce

Please sign in to comment.