Update README.md

README.md

@@ -152,9 +152,11 @@ It is a "large" versions of Conformer-CTC (around 120M parameters) model.
 This model can be efficiently (best latency and throughput) deployed with [NVIDIA Riva](https://developer.nvidia.com/riva), a GPU-accelerated speech AI SDK, on-premises, embedded, on the edge or with any cloud provider.
 
 Additionally, with RIVA you get:
+* Customization to achieve best WER 
+  * Ability to boost specific words (e.g. brand and product names)
+  * External language model, punctuation and captialization
+  * Conformer checkpoints trained on proprietary data
 * Streaming speech recognition mode
-* Ability to boost specific words (e.g. brand and product names)
-* Conformer checkpoints trained on proprietary data
 
 [Live Riva demo](https://developer.nvidia.com/riva#demos)  
 
@@ -189,9 +191,7 @@ asr_model.transcribe(['2086-149220-0033.wav'])
 ### Transcribing many audio files
 
 ```shell
-python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py \
- pretrained_name="nvidia/stt_en_conformer_ctc_large" \
- audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"
+python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py \n pretrained_name="nvidia/stt_en_conformer_ctc_large" \n audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"
 ```
 
 ### Input