Update README.md
Browse files
README.md
CHANGED
|
@@ -11,7 +11,7 @@ base_model: meta-llama/Llama-3.2-3B-Instruct
|
|
| 11 |
<p align="center">
|
| 12 |
<a href="https://docs.inference.net/use-cases/json-extraction"><strong>Documentation</strong></a> 路
|
| 13 |
<a href="https://inference.net/models/schematron-3b"><strong>Serverless API</strong></a> 路
|
| 14 |
-
<a href="https://
|
| 15 |
</p>
|
| 16 |
|
| 17 |
<br>
|
|
@@ -68,7 +68,7 @@ We evaluated Schematron's real-world impact on LLM factuality using SimpleQA.
|
|
| 68 |
3. **Structured Extraction**: Schematron extracts JSON data from retrieved pages using the schema
|
| 69 |
4. **Answer Synthesis**: Primary LLM produces final answer from structured data
|
| 70 |
|
| 71 |
-
![image/png](https://cdn-uploads.huggingface.co/production/uploads/6626a246891c75742bd19aaf/
|
| 72 |
|
| 73 |
**Key findings:**
|
| 74 |
- Web search paired with JSON extraction improves factuality: Adding Schematron with web retrieval improves GPT-5 Nano's accuracy from 8.54% to 82.87%鈥攏early a 10x improvement
|
|
|
|
| 11 |
<p align="center">
|
| 12 |
<a href="https://docs.inference.net/use-cases/json-extraction"><strong>Documentation</strong></a> 路
|
| 13 |
<a href="https://inference.net/models/schematron-3b"><strong>Serverless API</strong></a> 路
|
| 14 |
+
<a href="https://inference.net/blog/Schematron"><strong>Announcement blog</strong></a>
|
| 15 |
</p>
|
| 16 |
|
| 17 |
<br>
|
|
|
|
| 68 |
3. **Structured Extraction**: Schematron extracts JSON data from retrieved pages using the schema
|
| 69 |
4. **Answer Synthesis**: Primary LLM produces final answer from structured data
|
| 70 |
|
| 71 |
+
|
| 72 |
|
| 73 |
**Key findings:**
|
| 74 |
- Web search paired with JSON extraction improves factuality: Adding Schematron with web retrieval improves GPT-5 Nano's accuracy from 8.54% to 82.87%鈥攏early a 10x improvement
|