---
base_model: google/datagemma-rag-27b-it
inference: false
license: gemma
library_name: transformers
pipeline_tag: text-generation
model_creator: Google
model_name: datagemma-rag-27b-it
quantized_by: Second State Inc.
tags:
- conversational
---
# Datagemma-rag-27b-it-GGUF
## Original Model
[google/datagemma-rag-27b-it](https://huggingface.co/google/datagemma-rag-27b-it)
## Run with LlamaEdge
- LlamaEdge version: [v0.14.3](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.14.3) and above
- Prompt template
- Prompt type: `gemma-instruct`
- Prompt string
```text
user
{user_message}
model
{model_message}model
```
- Context size: `8192`
- Run as LlamaEdge service
```bash
wasmedge --dir .:. --nn-preload default:GGML:AUTO:datagemma-rag-27b-it-Q5_K_M.gguf \
llama-api-server.wasm \
--prompt-template gemma-instruct \
--ctx-size 8192 \
--model-name gemma-2-27b
```
- Run as LlamaEdge command app
```bash
wasmedge --dir .:. \
--nn-preload default:GGML:AUTO:datagemma-rag-27b-it-Q5_K_M.gguf \
llama-chat.wasm \
--prompt-template gemma-instruct \
--ctx-size 8192
```
## Quantized GGUF Models
| Name | Quant method | Bits | Size | Use case |
| ---- | ---- | ---- | ---- | ----- |
| [datagemma-rag-27b-it-Q2_K.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q2_K.gguf) | Q2_K | 2 | 10.4 GB| smallest, significant quality loss - not recommended for most purposes |
| [datagemma-rag-27b-it-Q3_K_L.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q3_K_L.gguf) | Q3_K_L | 3 | 14.5 GB| small, substantial quality loss |
| [datagemma-rag-27b-it-Q3_K_M.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q3_K_M.gguf) | Q3_K_M | 3 | 13.4 GB| very small, high quality loss |
| [datagemma-rag-27b-it-Q3_K_S.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q3_K_S.gguf) | Q3_K_S | 3 | 12.2 GB| very small, high quality loss |
| [datagemma-rag-27b-it-Q4_0.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q4_0.gguf) | Q4_0 | 4 | 15.6 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
| [datagemma-rag-27b-it-Q4_K_M.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q4_K_M.gguf) | Q4_K_M | 4 | 16.6 GB| medium, balanced quality - recommended |
| [datagemma-rag-27b-it-Q4_K_S.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q4_K_S.gguf) | Q4_K_S | 4 | 15.7 GB| small, greater quality loss |
| [datagemma-rag-27b-it-Q5_0.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q5_0.gguf) | Q5_0 | 5 | 18.9 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
| [datagemma-rag-27b-it-Q5_K_M.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q5_K_M.gguf) | Q5_K_M | 5 | 19.4 GB| large, very low quality loss - recommended |
| [datagemma-rag-27b-it-Q5_K_S.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q5_K_S.gguf) | Q5_K_S | 5 | 18.9 GB| large, low quality loss - recommended |
| [datagemma-rag-27b-it-Q6_K.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q6_K.gguf) | Q6_K | 6 | 22.3 GB| very large, extremely low quality loss |
| [datagemma-rag-27b-it-Q8_0.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-Q8_0.gguf) | Q8_0 | 8 | 28.9 GB| very large, extremely low quality loss - not recommended |
| [datagemma-rag-27b-it-f16-00001-of-00002.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-f16-00001-of-00002.gguf) | f16 | 16 | 29.9 GB| |
| [datagemma-rag-27b-it-f16-00002-of-00002.gguf](https://huggingface.co/second-state/datagemma-rag-27b-it-GGUF/blob/main/datagemma-rag-27b-it-f16-00002-of-00002.gguf) | f16 | 16 | 24.6 GB| |
*Quantized with llama.cpp b3664*