Hugging Face's logo

QuantFactory
/
Mistral-Gutenberg-Doppel-7B-FFT-GGUF

Transformers
GGUF
Inference Endpoints
conversational
Model card Files Community
Mistral-Gutenberg-Doppel-7B-FFT-GGUF
File size: 1,384 Bytes 
0d51968
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
33
34
 

---

license: apache-2.0
library_name: transformers
base_model:
- mistralai/Mistral-7B-Instruct-v0.2
datasets:
- jondurbin/gutenberg-dpo-v0.1
- nbeerbower/gutenberg2-dpo

---

[![QuantFactory Banner](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeiuCm7c8lEwEJuRey9kiVZsRn2W-b4pWlu3-X534V3YmVuVc2ZL-NXg2RkzSOOS2JXGHutDuyyNAUtdJI65jGTo8jT9Y99tMi4H4MqL44Uc5QKG77B0d6-JfIkZHFaUA71-RtjyYZWVIhqsNZcx8-OMaA?key=xt3VSDoCbmTY7o-cwwOFwQ)](https://hf.co/QuantFactory)


# QuantFactory/Mistral-Gutenberg-Doppel-7B-FFT-GGUF
This is quantized version of [nbeerbower/Mistral-Gutenberg-Doppel-7B-FFT](https://huggingface.co/nbeerbower/Mistral-Gutenberg-Doppel-7B-FFT) created using llama.cpp

# Original Model Card


![image/png](https://huggingface.co/nbeerbower/Mistral-Small-Gutenberg-Doppel-22B/resolve/main/doppel-header?download=true)

# Mistral-Gutenberg-Doppel-7B-FFT

[mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) finetuned on [jondurbin/gutenberg-dpo-v0.1](https://huggingface.co/datasets/jondurbin/gutenberg-dpo-v0.1) and [nbeerbower/gutenberg2-dpo](https://huggingface.co/datasets/nbeerbower/gutenberg2-dpo).

This is a full finetune rather than my usual QLoRA tunes. Mostly for learning purposes.

### Method

[ORPO tuned](https://mlabonne.github.io/blog/posts/2024-04-19_Fine_tune_Llama_3_with_ORPO.html) with 4x A100 for 2 epochs.