You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/source/en/perf_infer_gpu_multi.md
+4Lines changed: 4 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -306,3 +306,7 @@ The most important part of DTensor is the `placement` attribute because it tells
306
306
```
307
307
308
308
-`Partial()`- Indicates a tensor is pending a reduction operation (not typically relevant for usage in Transformers).
309
+
310
+
## Resources
311
+
312
+
Read the [Tensor Parallelism (TP) in Transformers: 5 Minutes to Understand](https://huggingface.co/blog/qgallouedec/tp) blog post for a quick overview of tensor parallelism and learn how column and row parallel setups differ.
0 commit comments