Update model card

Signed-off-by: Jinhan <jinhanw@nvidia.com>

README.md

@@ -199,17 +199,34 @@ img {
 </style>
 
 <p align="center">
-  <a href="#model-architecture"><img src="https://img.shields.io/badge/Model_Arch-FastConformer--CacheAware--RNNT-lightgrey#model-badge" alt="Model architecture"/></a>
+  <a href="#model-architecture">
+    <img src="https://img.shields.io/badge/Model_Arch-FastConformer--CacheAware--RNNT-76b900?style=flat#model-badge" alt="Model architecture"/>
+  </a>
   &nbsp;
-  <a href="#model-architecture"><img src="https://img.shields.io/badge/Params-600M-lightgrey#model-badge" alt="Model size"/></a>
+  <a href="#model-architecture">
+    <img src="https://img.shields.io/badge/Params-600M-76b900?style=flat#model-badge" alt="Model size"/>
+  </a>
   &nbsp;
-  <a href="#supported-languages"><img src="https://img.shields.io/badge/Language-Multilingual-lightgrey#model-badge" alt="Language"/></a>
+  <a href="#supported-languages">
+    <img src="https://img.shields.io/badge/Language-Multilingual-76b900?style=flat#model-badge" alt="Language"/>
+  </a>
+  <a href="https://developer.nvidia.com/nemotron" target="_blank" style="margin: 2px;">
+    <img alt="Homepage" src="https://img.shields.io/badge/🏠Nemotron Developer Page-Learn More Here!-536af5?color=76B900&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
+  </a>
+<a href="https://discord.gg/9xpKQtVvrk" target="_blank" style="margin: 2px;">
+    <img alt="Discord" src="https://img.shields.io/badge/Discord-NVIDIA%20AI%20Developer-7289da?logo=discord&logoColor=white&color=7289da" style="display: inline-block; vertical-align: middle;"/>
+  </a>
 </p>
 
 <p align="center">
   <img src="model_overview.png" alt="Nemotron 3.5 ASR overview: multilingual audio across 40 language-locales is transcribed by a cache-aware FastConformer-RNNT model with language-ID prompting into punctuated text with an automatic language tag" width="900"/>
 </p>
 
+<p align="center">
+  <img src="latency_vs_parallel.png" alt="Median final-token latency vs number of parallel requests on a single H100, Nemotron vs Parakeet RNNT across chunk sizes" width="900"/>
+</p>
+
+
 > [!Note]
 > This model is the multilingual extension of [nvidia/nemotron-speech-streaming-en-0.6b](https://huggingface.co/nvidia/nemotron-speech-streaming-en-0.6b), adding language-ID prompt conditioning to support transcription across **40 language-locales** from a single model.
 
@@ -221,34 +238,10 @@ It was trained on a massive ASR dataset and is engineered to perform across dive
 
 This model is ready for commercial use.
 
----
-
-## License/Terms of Use
-
-Governing Terms: Use of the model is governed by the [OpenMDW-1.1](https://openmdw.ai/license/1-1/) license.
-
-## Deployment Geography
-
-Global
-
-## Use Case
-
-This model is for transcription of multilingual audio.
-
 ## Release Date
 
 - Hugging Face [06/04/2026] via https://huggingface.co/nvidia/nemotron-3.5-asr-streaming-0.6b
 
-## References
-
-<a id="ref-1"></a>[1] [Stateful Conformer with Cache-based Inference for Streaming Automatic Speech Recognition](https://arxiv.org/abs/2312.17279)
-
-<a id="ref-2"></a>[2] [Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition](https://arxiv.org/abs/2305.05084)
-
-<a id="ref-3"></a>[3] [NVIDIA Granary](https://huggingface.co/datasets/nvidia/Granary)
-
-<a id="ref-4"></a>[4] [NVIDIA NeMo Framework](https://github.com/NVIDIA/NeMo)
-
 ## Why Choose Nemotron 3.5 ASR?
 
 - 🌍 **Single Multilingual Model:** Transcribes 40 language-locales from one model through language-ID prompt conditioning, with optional automatic language detection.
@@ -257,6 +250,8 @@ This model is for transcription of multilingual audio.
 - 🎛️ **Dynamic Runtime Flexibility:** Choose the optimal operating point on the latency-accuracy Pareto curve at inference time. No re-training is required to adjust for different use-case requirements.
 - 📝 **Punctuation & Capitalization:** Built-in support for punctuation and capitalization in output text.
 
+- 🔧 **Fine-tuning** Check our [blog post](https://huggingface.co/blog/nvidia/fine-tuning-nemotron-35-asr) of **how to fine-tune Nemotron 3.5 ASR to improve these languages**, including before/after results.
+
 ---
 
 ## Supported Languages
@@ -294,7 +289,7 @@ This model consists of a cache-aware streaming Parakeet (FastConformer) encoder
 
 The language-ID prompt is fused with the acoustic representation as follows:
 
-- **FastConformer encoder** processes audio into an acoustic embedding of shape (D=512, T).
+- **FastConformer encoder** processes audio into an acoustic embedding of shape (D=1024, T).
 - **Language Encoding** expands a 128-dim one-hot language vector across the time axis → (K=128, T), broadcasting the language identity to every frame.
 - **Concatenation** along the feature axis → fused tensor (D + K, T).
 - **Projection layer** maps the fused features to the RNNT decoder.
@@ -359,7 +354,7 @@ Also, check out the following NVIDIA speech models:
 
 ## NVIDIA NeMo
 
-To train, fine-tune or perform inference with this model, you will need to install [NVIDIA NeMo](https://github.com/NVIDIA/NeMo) [\[4\]](#ref-4). We recommend you install it after you've installed Cython and latest PyTorch version.
+To train, fine-tune or perform inference with this model, you will need to install [NVIDIA NeMo](https://github.com/NVIDIA/NeMo) [\[4\]](#ref-4). We recommend you install it after you've installed python >3.11, Cython and latest PyTorch version.
 
 ```bash
 apt-get update && apt-get install -y libsndfile1 ffmpeg
@@ -579,8 +574,29 @@ _Languages are ordered by accuracy (lowest WER first)._
 
 These **8 language-locales** are recognized by the tokenizer but are not tuned for production transcription out of the box: **Greek (el-GR), Hebrew (he-IL), Lithuanian (lt-LT), Slovenian (sl-SI), Latvian (lv-LV), Maltese (mt-MT), Thai (th-TH), and Norwegian Nynorsk (nn-NO)**. Fine-tuning on in-domain data is recommended to bring them to production quality.
 
-Check our [blog post](https://huggingface.co/blog/nvidia/fine-tuning-nemotron-35-asr) of **how to fine-tune Nemotron 3.5 ASR to improve these languages**, including before/after results.
+---
+
+## License/Terms of Use
+
+Governing Terms: Use of the model is governed by the [OpenMDW-1.1](https://openmdw.ai/license/1-1/) license.
+
+## Deployment Geography
+
+Global
+
+## Use Case
+
+This model is for transcription of multilingual audio.
 
+## References
+
+<a id="ref-1"></a>[1] [Stateful Conformer with Cache-based Inference for Streaming Automatic Speech Recognition](https://arxiv.org/abs/2312.17279)
+
+<a id="ref-2"></a>[2] [Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition](https://arxiv.org/abs/2305.05084)
+
+<a id="ref-3"></a>[3] [NVIDIA Granary](https://huggingface.co/datasets/nvidia/Granary)
+
+<a id="ref-4"></a>[4] [NVIDIA NeMo Framework](https://github.com/NVIDIA/NeMo)
 
 ---