add links from fsdp and deepspeed docs. fix deepspeed imports

huggingface · May 1, 2024 · 3b1948f · 3b1948f
1 parent 7183706
commit 3b1948f
Show file tree

Hide file tree

Showing 3 changed files with 19 additions and 5 deletions.
diff --git a/docs/source/package_reference/deepspeed.md b/docs/source/package_reference/deepspeed.md
@@ -17,12 +17,12 @@ rendered properly in your Markdown viewer.
 
 [[autodoc]] utils.DeepSpeedPlugin
 
-[[autodoc]] utils.DummyOptim
+[[autodoc]] utils.deepspeed.DummyOptim
 
-[[autodoc]] utils.DummyScheduler
+[[autodoc]] utils.deepspeed.DummyScheduler
 
-[[autodoc]] utils.DeepSpeedEngineWrapper
+[[autodoc]] utils.deepspeed.DeepSpeedEngineWrapper
 
-[[autodoc]] utils.DeepSpeedOptimizerWrapper
+[[autodoc]] utils.deepspeed.DeepSpeedOptimizerWrapper
 
-[[autodoc]] utils.DeepSpeedSchedulerWrapper
+[[autodoc]] utils.deepspeed.DeepSpeedSchedulerWrapper
diff --git a/docs/source/usage_guides/deepspeed.md b/docs/source/usage_guides/deepspeed.md
@@ -729,3 +729,10 @@ Papers:
 
 Finally, please, remember that 🤗 `Accelerate` only integrates DeepSpeed, therefore if you
 have any problems or questions with regards to DeepSpeed usage, please, file an issue with [DeepSpeed GitHub](https://github.com/microsoft/DeepSpeed/issues).
+
+
+<Tip>
+
+    For those interested in the similarities and differences between FSDP and DeepSpeed, please check out the [concept guide here](../concept_guides/fsdp_and_deepspeed.md)!
+
+</Tip>
diff --git a/docs/source/usage_guides/fsdp.md b/docs/source/usage_guides/fsdp.md
@@ -175,3 +175,10 @@ You can then pass `state` into the `save_pretrained` method.  There are several
 
 For more control, users can leverage the `FullyShardedDataParallelPlugin`. After creating an instance of this class, users can pass it to the Accelerator class instantiation.
 For more information on these options, please refer to the PyTorch [FullyShardedDataParallel](https://github.com/pytorch/pytorch/blob/0df2e863fbd5993a7b9e652910792bd21a516ff3/torch/distributed/fsdp/fully_sharded_data_parallel.py#L236) code.
+
+
+<Tip>
+
+    For those interested in the similarities and differences between FSDP and DeepSpeed, please check out the [concept guide here](../concept_guides/fsdp_and_deepspeed.md)!
+
+</Tip>