File size: 3,590 Bytes
740d59f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- moe
- merge
- mergekit
- vicgalle/CarbonBeagle-11B
- Sao10K/Fimbulvetr-10.7B-v1
- bn22/Nous-Hermes-2-SOLAR-10.7B-MISALIGNED
- Yhyu13/LMCocktail-10.7B-v1
---

# Umbra-v2-MoE-4x10.7

Umbra-v2-MoE-4x10.7 is a Mixure of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [vicgalle/CarbonBeagle-11B](https://huggingface.co./vicgalle/CarbonBeagle-11B)
* [Sao10K/Fimbulvetr-10.7B-v1](https://huggingface.co./Sao10K/Fimbulvetr-10.7B-v1)
* [bn22/Nous-Hermes-2-SOLAR-10.7B-MISALIGNED](https://huggingface.co./bn22/Nous-Hermes-2-SOLAR-10.7B-MISALIGNED)
* [Yhyu13/LMCocktail-10.7B-v1](https://huggingface.co./Yhyu13/LMCocktail-10.7B-v1)

## 🧩 Configuration

```yamlbase_model: vicgalle/CarbonBeagle-11B
gate_mode: hidden
dtype: bfloat16

experts:
  - source_model: vicgalle/CarbonBeagle-11B
    positive_prompts:
    - "versatile"
    - "adaptive"
    - "comprehensive"
    - "integrated"
    - "balanced"
    - "all-rounder"
    - "flexible"
    - "wide-ranging"
    - "multi-disciplinary"
    - "holistic"
    - "innovative"
    - "eclectic"
    - "resourceful"
    - "dynamic"
    - "robust"

    negative_prompts:
    - "narrow"
    - "specialized"
    - "limited"
    - "focused"

  - source_model: Sao10K/Fimbulvetr-10.7B-v1
    positive_prompts:
    - "creative"
    - "storytelling"
    - "expressive"
    - "imaginative"
    - "engaging"
    - "verbose"
    - "narrative"
    - "descriptive"
    - "elaborate"
    - "fictional"
    - "artistic"
    - "vivid"
    - "colorful"
    - "fantastical"
    - "lyrical"

    negative_prompts:
    - "sorry"
    - "I cannot"
    - "factual"
    - "concise"
    - "straightforward"
    - "objective"
    - "dry"

  - source_model: bn22/Nous-Hermes-2-SOLAR-10.7B-MISALIGNED
    positive_prompts:
    - "intelligent"
    - "analytical"
    - "accurate"
    - "knowledgeable"
    - "logical"
    - "data-driven"
    - "scientific"
    - "rational"
    - "precise"
    - "methodical"
    - "empirical"
    - "systematic"
    - "efficient"
    - "scholarly"
    - "statistical"
    - "calculate"
    - "compute"
    - "solve"
    - "work"
    - "python"
    - "javascript"
    - "programming"
    - "algorithm"
    - "tell me"
    - "assistant"

    negative_prompts:
    - "creative"
    - "imaginative"
    - "abstract"
    - "emotional"
    - "artistic"
    - "speculative"

  - source_model: Yhyu13/LMCocktail-10.7B-v1
    positive_prompts:
    - "instructive"
    - "verbose"
    - "descriptive"
    - "clear"
    - "detailed"
    - "informative"
    - "explanatory"
    - "elucidative"
    - "articulate"
    - "comprehensive"
    - "educational"
    - "thorough"
    - "specific"
    - "clarifying"
    - "structured"

    negative_prompts:
    - "concise"
    - "vague"```

## 💻 Usage

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

from transformers import AutoTokenizer
import transformers
import torch

model = "Steelskull/Umbra-v2-MoE-4x10.7"

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"])
```