Add NVIDIA BERT (#9949)

Signed-off-by: Pablo Ribalta Lorenzo <pribalta@nvidia.com>

community/README.md

@@ -45,6 +45,8 @@ This repository provides a curated list of the GitHub repositories with machine
 | [Transformer-LT (MLPerf)](https://github.com/IntelAI/models/tree/master/benchmarks/language_translation/tensorflow/transformer_mlperf) | [Attention Is All You Need](https://arxiv.org/pdf/1706.03762) | • FP32 Training | [Intel](https://github.com/IntelAI) |
 | [Transformer-LT (Official)](https://github.com/IntelAI/models/tree/master/benchmarks/language_translation/tensorflow/transformer_lt_official) | [Attention Is All You Need](https://arxiv.org/pdf/1706.03762) | • FP32 Inference | [Intel](https://github.com/IntelAI) |
 | [ELECTRA](https://github.com/NVIDIA/DeepLearningExamples/tree/master/TensorFlow2/LanguageModeling/ELECTRA) | [ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators](https://openreview.net/forum?id=r1xMH1BtvB) | • Automatic Mixed Precision<br/>• Multi-GPU training support with Horovod<br/>• Multi-node training on a Pyxis/Enroot Slurm cluster | [NVIDIA](https://github.com/NVIDIA) |
+| [BERT](https://github.com/NVIDIA/DeepLearningExamples/tree/master/TensorFlow2/LanguageModeling/BERT) | [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://arxiv.org/pdf/1810.04805) | • Horovod Multi-GPU<br/>• Multi-node with Horovod and Pyxis/Enroot Slurm cluster<br/>• XLA<br/>• Automatic mixed precision<br/>• LAMB | [NVIDIA](https://github.com/NVIDIA) |
+
 
 ## Recommendation Systems