File size: 3,396 Bytes
0d44b67
 
 
 
 
 
 
 
 
c922573
 
 
 
 
 
 
0d44b67
c922573
 
 
0d44b67
 
 
 
 
 
 
 
 
 
 
 
 
 
5546501
0d44b67
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
---
base_model: []
library_name: transformers
tags:
- mergekit
- merge

---

---
### (GGUF) Thanks:

#### mradermacher

- **GGUF:** **[mradermacher/LexiMaid-L3-8B-GGUF](https://huggingface.co./mradermacher/LexiMaid-L3-8B-GGUF)**
- **imatrix GGUF:** **[mradermacher/LexiMaid-L3-8B-i1-GGUF](https://huggingface.co./mradermacher/LexiMaid-L3-8B-i1-GGUF)**

---

This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).

---

**Leximaid** - This experimental model was created based on discussions on Reddit, where models such as **[Lumimaid-v0.2-8B](https://huggingface.co./NeverSleep/Lumimaid-v0.2-8B)**, **[Dusk_Rainbow](https://huggingface.co./SicariusSicariiStuff/Dusk_Rainbow)**, **[Stheno-v3.4](https://huggingface.co./Sao10K/Llama-3.1-8B-Stheno-v3.4)**, **[Celeste-V1.5](https://huggingface.co./nothingiisreal/L3.1-8B-Celeste-V1.5)**, and **[Lexi-Uncensored-V2](https://huggingface.co./Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2)** were mentioned — each with its own strengths and weaknesses. Some of them received criticism, which inspired me to create a new synthesis.

To merge the models, I used the **dare ties** method, first combining **[Lumimaid-v0.2-8B](https://huggingface.co./NeverSleep/Lumimaid-v0.2-8B)**, **[Dusk_Rainbow](https://huggingface.co./SicariusSicariiStuff/Dusk_Rainbow)**, and **[Stheno-v3.4](https://huggingface.co./Sao10K/Llama-3.1-8B-Stheno-v3.4)**, and then mixing the resulting model with **[Celeste-V1.5](https://huggingface.co./nothingiisreal/L3.1-8B-Celeste-V1.5)** and **[Lexi-Uncensored-V2](https://huggingface.co./Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2)**. **Leximaid** is focused on roleplay and creative storytelling.

---


## Merge Details

### Merge Method

This model was merged using the [DARE](https://arxiv.org/abs/2311.03099) [TIES](https://arxiv.org/abs/2306.01708).

### Models Merged

The following models were included in the merge:

**[Lumimaid-v0.2-8B](https://huggingface.co./NeverSleep/Lumimaid-v0.2-8B)**
**[Dusk_Rainbow](https://huggingface.co./SicariusSicariiStuff/Dusk_Rainbow)**
**[Stheno-v3.4](https://huggingface.co./Sao10K/Llama-3.1-8B-Stheno-v3.4)**


**[Celeste-V1.5](https://huggingface.co./nothingiisreal/L3.1-8B-Celeste-V1.5)**
**[Lexi-Uncensored-V2](https://huggingface.co./Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2)**

## Configuration

The following YAML configuration was used to produce Leximaid:

```yaml
models:
  - model: Arkana08/Maxi-Fail-L3-8b
    parameters:
      weight: 0.4
      density: 0.7
  - model: nothingiisreal/L3.1-8B-Celeste-V1.5
    parameters:
      weight: 0.3
      density: 0.75
  - model: Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2
    parameters:
      weight: 0.3
      density: 0.65
merge_method: dare_ties
base_model: Arkana08/Maxi-Fail-L3-8b
parameters:
  int8_mask: true
dtype: bfloat16
```

## Credits
Thanks to the creators of the models:
- **[NeverSleep/Lumimaid-v0.2-8B](https://huggingface.co./NeverSleep/Lumimaid-v0.2-8B)**
- **[SicariusSicariiStuff/Dusk_Rainbow](https://huggingface.co./SicariusSicariiStuff/Dusk_Rainbow)**
- **[Sao10K/Llama-3.1-8B-Stheno-v3.4](https://huggingface.co./Sao10K/Llama-3.1-8B-Stheno-v3.4)**
- **[nothingiisreal/L3.1-8B-Celeste-V1.5](https://huggingface.co./nothingiisreal/L3.1-8B-Celeste-V1.5)**
- **[Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2](https://huggingface.co./Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2)**