Upload README.md
Browse files
README.md
ADDED
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model:
|
3 |
+
- Epiculous/Fett-uccine-7B
|
4 |
+
- eren23/ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v2
|
5 |
+
- OpenPipe/mistral-ft-optimized-1227
|
6 |
+
- ChaoticNeutrals/Eris_7B
|
7 |
+
library_name: transformers
|
8 |
+
license: cc-by-nc-4.0
|
9 |
+
tags:
|
10 |
+
- mergekit
|
11 |
+
- merge
|
12 |
+
language:
|
13 |
+
- en
|
14 |
+
pipeline_tag: text-generation
|
15 |
+
---
|
16 |
+
# merge
|
17 |
+
|
18 |
+
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
19 |
+
|
20 |
+
## Merge Details
|
21 |
+
* This model is an attempt at making a smart rp model with the finesse of [Epiculous/Fett-uccine-7B](https://huggingface.co/Epiculous/Fett-uccine-7B).
|
22 |
+
* From limited testing i've found it to be my favourite of my personal 7B models. It stays pretty coherent at 8k+ ctx.
|
23 |
+
* I like to use "Alpaca" format with "Universal-Light" for longer messages. Switching to ChatML causes the messages to be much shorter? I haven't a clue why but sometimes it's nice.
|
24 |
+
* It doesn't seem to show many issues but i'd be willing to try to fix any problems or bugs as it shows some potential.
|
25 |
+
### Merge Method
|
26 |
+
|
27 |
+
This model was merged using the [DARE](https://arxiv.org/abs/2311.03099) [TIES](https://arxiv.org/abs/2306.01708) merge method using [OpenPipe/mistral-ft-optimized-1227](https://huggingface.co/OpenPipe/mistral-ft-optimized-1227) as a base.
|
28 |
+
|
29 |
+
### Models Merged
|
30 |
+
|
31 |
+
The following models were included in the merge:
|
32 |
+
* [Epiculous/Fett-uccine-7B](https://huggingface.co/Epiculous/Fett-uccine-7B)
|
33 |
+
* [eren23/ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v2](https://huggingface.co/eren23/ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v2)
|
34 |
+
* [ChaoticNeutrals/Eris_7B](https://huggingface.co/ChaoticNeutrals/Eris_7B)
|
35 |
+
|
36 |
+
### Configuration
|
37 |
+
|
38 |
+
The following YAML configuration was used to produce this model:
|
39 |
+
|
40 |
+
```yaml
|
41 |
+
models:
|
42 |
+
- model: OpenPipe/mistral-ft-optimized-1227
|
43 |
+
# No parameters necessary for base model
|
44 |
+
- model: Epiculous/Fett-uccine-7B
|
45 |
+
parameters:
|
46 |
+
density: 0.53
|
47 |
+
weight: 0.4
|
48 |
+
- model: eren23/ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v2
|
49 |
+
parameters:
|
50 |
+
density: 0.53
|
51 |
+
weight: 0.35
|
52 |
+
- model: ChaoticNeutrals/Eris_7B
|
53 |
+
parameters:
|
54 |
+
density: 0.53
|
55 |
+
weight: 0.25
|
56 |
+
merge_method: dare_ties
|
57 |
+
base_model: OpenPipe/mistral-ft-optimized-1227
|
58 |
+
parameters:
|
59 |
+
int8_mask: true
|
60 |
+
dtype: bfloat16
|
61 |
+
|
62 |
+
```
|