Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -37,7 +37,7 @@ Sarus executes an internal **Cognitive Reasoning Monologue** (`[THINK] ... [/THI
|
|
| 37 |
|
| 38 |
## 🌟 Key Features
|
| 39 |
|
| 40 |
-
* **🧠Cognitive
|
| 41 |
* **🇮🇳 Bilingual Fluency**: Native support for English, Hindi, and romanized Hinglish.
|
| 42 |
* **âš¡ High Efficiency**: Optimized for low-latency edge and server deployment using Grouped Query Attention (GQA).
|
| 43 |
* **💬 Conversational**: Multi-turn dialogue support with standard chat formatting.
|
|
@@ -65,9 +65,6 @@ Sarus executes an internal **Cognitive Reasoning Monologue** (`[THINK] ... [/THI
|
|
| 65 |
<|user|>
|
| 66 |
{user_message}<|endofturn|>
|
| 67 |
<|assistant|>
|
| 68 |
-
[THINK]
|
| 69 |
-
{internal_reasoning}
|
| 70 |
-
[/THINK]
|
| 71 |
{final_answer}<|endofturn|>
|
| 72 |
```
|
| 73 |
|
|
|
|
| 37 |
|
| 38 |
## 🌟 Key Features
|
| 39 |
|
| 40 |
+
* **🧠Cognitive Intelligence**: Structured, direct-response reasoning trained on STEM, facts, and logic.
|
| 41 |
* **🇮🇳 Bilingual Fluency**: Native support for English, Hindi, and romanized Hinglish.
|
| 42 |
* **âš¡ High Efficiency**: Optimized for low-latency edge and server deployment using Grouped Query Attention (GQA).
|
| 43 |
* **💬 Conversational**: Multi-turn dialogue support with standard chat formatting.
|
|
|
|
| 65 |
<|user|>
|
| 66 |
{user_message}<|endofturn|>
|
| 67 |
<|assistant|>
|
|
|
|
|
|
|
|
|
|
| 68 |
{final_answer}<|endofturn|>
|
| 69 |
```
|
| 70 |
|