File size: 668 Bytes
61ec6c7
 
 
 
 
6eac43e
f55d3a5
d0cc92f
 
 
 
 
 
 
61ec6c7
d0cc92f
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
---
library_name: transformers
tags: []
---

# gemma2-mitra-it-int8
This is an 8int quantized version of gemma-2-mitra-it: https://huggingface.co./buddhist-nlp/gemma-2-mitra-it  
The quantization was done with llm compressor: https://github.com/vllm-project/llm-compressor  

The template for prompting the model is this: 
```
Please translate into <target_language>: <input_sentence> 🔽 Translation::
```
Line breaks in this model should be replaced with the '🔽' character before running the generation. '#' is used as a stop token.  
## Model Details
For details on how to run this please see the gemma2-9b repository: https://huggingface.co./google/gemma-2-9b