update doc for perf_train_cpu_many, add intel mpi introduction (huggingface#18576)

* update doc for perf_train_cpu_many, add mpi introduction

Signed-off-by: Wang, Yi A <[email protected]>

* Update docs/source/en/perf_train_cpu_many.mdx

Co-authored-by: Sylvain Gugger <[email protected]>

* Update docs/source/en/perf_train_cpu_many.mdx

Signed-off-by: Wang, Yi A <[email protected]>

Signed-off-by: Wang, Yi A <[email protected]>
Co-authored-by: Sylvain Gugger <[email protected]>

Author: 2 people

docs/source/en/perf_train_cpu_many.mdx

@@ -36,8 +36,22 @@ pip install oneccl_bind_pt=={pytorch_version} -f https://software.intel.com/ipex
 ```
 where `{pytorch_version}` should be your PyTorch version, for instance 1.12.0.
 Check more approaches for [oneccl_bind_pt installation](https://github.com/intel/torch-ccl).
+Versions of oneCCL and PyTorch must match.
 
-### Usage in Trainer
+## Intel® MPI library
+Use this standards-based MPI implementation to deliver flexible, efficient, scalable cluster messaging on Intel® architecture. This component is part of the Intel® oneAPI HPC Toolkit.
+It can be installed via [MPI](https://www.intel.com/content/www/us/en/developer/articles/tool/oneapi-standalone-components.html#mpi).
+
+Please set the environment by following command before using it.
+
+```
+source /opt/intel/oneapi/setvars.sh
+```
+
+The following "Usage in Trainer" takes mpirun in Intel® MPI library as an example.
+
+
+## Usage in Trainer
 To enable multi CPU distributed training in the Trainer with the ccl backend, users should add **`--xpu_backend ccl`** in the command arguments.
 
 Let's see an example with the [question-answering example](https://github.com/huggingface/transformers/tree/main/examples/pytorch/question-answering)