File size: 3,813 Bytes
fc804ad
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
---
license: apache-2.0
tags:
- moe
- frankenmoe
- merge
- mergekit
- lazymergekit
- Himitsui/Kaiju-11B
- Sao10K/Fimbulvetr-11B-v2
- decapoda-research/Antares-11b-v2
- beberik/Nyxene-v3-11B
base_model:
- Himitsui/Kaiju-11B
- Sao10K/Fimbulvetr-11B-v2
- decapoda-research/Antares-11b-v2
- beberik/Nyxene-v3-11B
---

# Umbra-v3-MoE-4x11b

Umbra-v3-MoE-4x11b is a Mixure of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [Himitsui/Kaiju-11B](https://huggingface.co/Himitsui/Kaiju-11B)
* [Sao10K/Fimbulvetr-11B-v2](https://huggingface.co/Sao10K/Fimbulvetr-11B-v2)
* [decapoda-research/Antares-11b-v2](https://huggingface.co/decapoda-research/Antares-11b-v2)
* [beberik/Nyxene-v3-11B](https://huggingface.co/beberik/Nyxene-v3-11B)

## 🧩 Configuration

```yaml
base_model: vicgalle/CarbonBeagle-11B-truthy
gate_mode: hidden
dtype: bfloat16
experts_per_token: 4
experts:
  - source_model: Himitsui/Kaiju-11B
    positive_prompts:
      - "Imagine"
      - "Create"
      - "Envision"
      - "Fantasize"
      - "Invent"
      - "Narrate"
      - "Plot"
      - "Portray"
      - "Storytell"
      - "Visualize"
      - "Describe"
      - "Develop"
      - "Forge"
      - "Craft"
      - "Conceptualize"
      - "Dream"
      - "Concoct"
      - "Characterize"
    negative_prompts:
      - "Recite"
      - "Report"
      - "Summarize"
      - "Enumerate"
      - "List"
      - "Cite"

  - source_model: Sao10K/Fimbulvetr-11B-v2
    positive_prompts:
      - "Dramatize"
      - "Embody"
      - "Illustrate"
      - "Perform"
      - "Roleplay"
      - "Simulate"
      - "Stage"
      - "Unfold"
      - "Weave"
      - "Design"
      - "Outline"
      - "Script"
      - "Sketch"
      - "Spin"
      - "Depict"
      - "Render"
      - "Fashion"
      - "Conceive"
    negative_prompts:
      - "Analyze"
      - "Critique"
      - "Dissect"
      - "Explain"
      - "Clarify"
      - "Interpret"

  - source_model: decapoda-research/Antares-11b-v2
    positive_prompts:
      - "Solve"
      - "Respond"
      - "Convey"
      - "Disclose"
      - "Expound"
      - "Narrate"
      - "Present"
      - "Reveal"
      - "Specify"
      - "Uncover"
      - "Decode"
      - "Examine"
      - "Report"
      - "Survey"
      - "Validate"
      - "Verify"
      - "Question"
      - "Query"
    negative_prompts:
      - "Divert"
      - "Obscure"
      - "Overstate"
      - "Undermine"
      - "Misinterpret"
      - "Skew"

  - source_model: beberik/Nyxene-v3-11B
    positive_prompts:
      - "Explain"
      - "Instruct"
      - "Clarify"
      - "Educate"
      - "Guide"
      - "Inform"
      - "Teach"
      - "Detail"
      - "Elaborate"
      - "Enlighten"
      - "Advise"
      - "Interpret"
      - "Analyze"
      - "Define"
      - "Demonstrate"
      - "Illustrate"
      - "Simplify"
      - "Summarize"
    negative_prompts:
      - "Speculate"
      - "Fabricate"
      - "Exaggerate"
      - "Mislead"
      - "Confuse"
      - "Distort"

```

## 💻 Usage

```python
!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Steelskull/Umbra-v3-MoE-4x11b"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
```