Update README.md

README.md

@@ -10,7 +10,7 @@ pinned: false
 <div align="center">
   <img src="https://netspresso-docs-imgs.s3.ap-northeast-2.amazonaws.com/imgs/banner/huggingfacenota.png" 
        alt="Nota AI Banner" 
-       style="width: 100%; height: 200px; object-fit: cover;" />
+       style="width: 100%; height: auto; max-width: 100%;" />
 </div>
 
 <br>
@@ -33,13 +33,14 @@ From our automated **optimization platform** to bespoke **AI solutions**, we ens
 > ## **World Best LLM (WBL) Project**
 > Nota AI participates in the **'World Best LLM' (WBL)** project, a key initiative by the South Korean government (NIPA) to develop global-tier foundation models. As a core optimization partner, we focus on compressing massive LLMs for practical deployment.
 >
-> ## **🔥 New Release: [Qwen3-30B-A3B-NotaMoEQuant-Int4](https://huggingface.co/nota-ai/Qwen3-30B-A3B-NotaMoEQuant-Int4)**
-> **4-bit Quantization for Mixture-of-Experts (MoE)**
+> ## **🔥 New Release: [Solar-Open-100B-NotaMoEQuant-Int4](https://huggingface.co/nota-ai/Solar-Open-100B-NotaMoEQuant-Int4)**
+> **Quantized Model for Upstage's Solar-Open-100B**
 >
-> This model demonstrates our proprietary **NotaMoEQuant** technology applied to the Qwen3-30B architecture.
-> * **Optimization Tech:** NotaMoEQuant (Int4 Quantization for Active Parameters).
-> * **Key Benefit:** Significantly reduces memory bandwidth requirements while maintaining reasoning capabilities of the 30B MoE model.
-> * **Target:** Efficient inference on consumer-grade GPUs and edge servers.
+> This model is optimized using our proprietary **NotaMoEQuant**, a specialized methodology for Mixture-of-Experts (MoE) architectures.
+> * **Why NotaMoEQuant:** Unlike conventional methods (e.g., AutoRound) that overlook expert routing changes during quantization, our approach directly resolves the resulting representational distortion, delivering superior benchmark accuracy.
+> * **Hardware Efficiency:** Reduces the GPU requirement for maximum context generation from **4x A100 (80GB) to 2x A100 (80GB)**, saving up to 50% on inference costs.
+> 
+> *Also available: [Solar-Open-100B-Nota-FP8](https://huggingface.co/nota-ai/Solar-Open-100B-Nota-FP8)*
 
 # 🚀 Our Core Business
 <table border="0" cellspacing="0" cellpadding="0" style="border: none; border-collapse: collapse; width: 100%;">