IMprove README page

Summary:
- Add Llama3.2 news to top-level README page

- Quick summary of what we did with Llama 3.2

Reviewed By: helunwencser

Differential Revision: D63429498

fbshipit-source-id: 18757500d0ab72c5c70b8dfaa9d6a629ce083316

Author: mergennachin

README.md

@@ -24,6 +24,9 @@ Check out the [Getting Started](https://pytorch.org/executorch/stable/getting-st
 
 Check out the examples of [Llama](./examples/models/llama2/README.md), [Llava](./examples/models/llava/README.md) and [other models](./examples/README.md) running on edge devices using ExecuTorch.
 
+
+**[UPDATE - 09/25]** We have added support for running [Llama 3.2 1B/3B](./examples/models/llama2/README.md) models via ExecuTorch.
+
 ## Feedback
 
 We welcome any feedback, suggestions, and bug reports from the community to help

examples/models/llama2/README.md

@@ -24,6 +24,10 @@ Please note that the models are subject to the [Llama 2 Acceptable Use Policy](h
 
 Since Llama 2 7B or Llama 3 8B model needs at least 4-bit quantization to fit even within some of the highend phones, results presented here correspond to 4-bit groupwise post-training quantized model.
 
+For Llama 3.2 1B/3B, we validated the models by running them in their original bf16 datatype and unquantized on both Android and iOS phones. The 3B version required high-end phones with larger RAMs to fit the model.
+
+Additionally, these models are sensitive to accuracy when regular PTQ quantization is applied, so we employed [SpinQuant](https://github.com/facebookresearch/SpinQuant/tree/main) to achieve a good balance between accuracy and performance.
+
 <table>
   <tr>
     <td>