Update README.md

README.md

@@ -16,8 +16,18 @@ pipeline_tag: question-answering
 
 ## Model Details
 
-This model is a mixture of experts (MoE) using the [RhuiDih/moetify](https://github.com/RhuiDih/moetify) library. It combines multiple domain-specific experts, LoRA adapters, and datasets, all available at [Moecule Ingredients]
-(https://huggingface.co/collections/davzoku/moecule-ingredients-67dac0e6210eb1d95abc6411).
+This model is a mixture of experts (MoE) using the [RhuiDih/moetify](https://github.com/RhuiDih/moetify) library. It combines multiple domain-specific experts, LoRA adapters, and datasets, all available at [Moecule Ingredients](https://huggingface.co/collections/davzoku/moecule-ingredients-67dac0e6210eb1d95abc6411).
+
+## Key Features
+
+- **Zero Additional Training:** Combine existing domain-specific / task-specific experts into a powerful MoE model without additional training!
+
+## System Requirements
+
+| Steps            | System Requirements    |
+| ---------------- | ---------------------- |
+| MoE Creation     | > 54.2 GB System RAM   |
+| Inference (fp16) | GPU with > 15.5GB VRAM |
 
 ## MoE Creation
 
@@ -39,7 +49,7 @@ To reproduce this model, run the following command:
        davzoku/stock_market_expert_3b
 ```
 
-### Model Parameters
+## Model Parameters
 
 ```shell
 INFO:root:Stem parameters: 1228581888
@@ -50,7 +60,7 @@ INFO:root:MOE total parameters : 8363609088
 INFO:root:MOE active parameters: 5985438720
 ```
 
-## Quick Start
+## Inference
 
 ```python
 # git clone moetify fork that fixes dependency issue
@@ -102,3 +112,4 @@ print(generated_text)
 
 - [Flexible and Effective Mixing of Large Language Models into a Mixture of Domain Experts](https://arxiv.org/abs/2408.17280v2)
 - [RhuiDih/moetify](https://github.com/RhuiDih/moetify)
+