File size: 1,104 Bytes
58b8613
 
 
e297626
 
cfd9ce2
58b8613
 
cfd9ce2
58b8613
cfd9ce2
58b8613
cfd9ce2
58b8613
 
 
 
cfd9ce2
 
66199a5
cfd9ce2
58b8613
 
66199a5
cfd9ce2
58b8613
 
66199a5
 
58b8613
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
---
license: apache-2.0
inference: false 
base_model: llmware/dragon-mistral-7b-v0
base_model_relation: quantized 
tags: [green, llmware-rag, p7, ov, emerald]
---

# dragon-mistral-ov

**dragon-mistral-ov** is a high-quality fact-based question-answering model, designed for retrieval augmented generation (RAG) with complex business documents, quantized and packaged in OpenVino int4 for AI PCs using Intel GPU, CPU and NPU.    

This model provides a good combination of quality and inference speed.     

### Model Description

- **Developed by:** llmware  
- **Model type:** mistral-0.1
- **Parameters:** 7 billion  
- **Quantization:** int4  
- **Model Parent:** [llmware/dragon-mistral-7b-v0](https://www.huggingface.co/llmware/dragon-mistral-7b-v0)    
- **Language(s) (NLP):** English  
- **License:** Apache 2.0  
- **Uses:** Fact-based question-answering, RAG  
- **RAG Benchmark Accuracy Score:** 96.5  


## Model Card Contact  
[llmware on github](https://www.github.com/llmware-ai/llmware)  
[llmware on hf](https://www.huggingface.co/llmware)  
[llmware website](https://www.llmware.ai)