Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,116 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- moe
|
5 |
+
- frankenmoe
|
6 |
+
- merge
|
7 |
+
- mergekit
|
8 |
+
- lazymergekit
|
9 |
+
- Gille/StrangeMerges_32-7B-slerp
|
10 |
+
- mlabonne/AlphaMonarch-7B
|
11 |
+
base_model:
|
12 |
+
- Gille/StrangeMerges_32-7B-slerp
|
13 |
+
- mlabonne/AlphaMonarch-7B
|
14 |
+
---
|
15 |
+
|
16 |
+
# MixtureofMerges-MoE-2x7b-v7
|
17 |
+
|
18 |
+
MixtureofMerges-MoE-2x7b-v7 is a Mixture of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
19 |
+
* [Gille/StrangeMerges_32-7B-slerp](https://huggingface.co/Gille/StrangeMerges_32-7B-slerp)
|
20 |
+
* [mlabonne/AlphaMonarch-7B](https://huggingface.co/mlabonne/AlphaMonarch-7B)
|
21 |
+
|
22 |
+
## 🧩 Configuration
|
23 |
+
|
24 |
+
```yaml
|
25 |
+
base_model: Gille/StrangeMerges_32-7B-slerp
|
26 |
+
gate_mode: hidden
|
27 |
+
dtype: bfloat16
|
28 |
+
experts:
|
29 |
+
- source_model: Gille/StrangeMerges_32-7B-slerp
|
30 |
+
positive_prompts:
|
31 |
+
- "Answer this question from the ARC (Argument Reasoning Comprehension)."
|
32 |
+
- "Use common sense and logical reasoning skills."
|
33 |
+
- "What assumptions does this argument rely on?"
|
34 |
+
- "Are these assumptions valid? Explain."
|
35 |
+
- "Analyze the logical structure of this argument. Identify the premises, conclusion, and any assumptions made"
|
36 |
+
- "Identify any potential counterarguments to this position. How might someone challenge the reasoning presented?"
|
37 |
+
- "Could this be explained in a different way? Provide an alternative explanation."
|
38 |
+
- "Identify any weaknesses in this argument."
|
39 |
+
- "Does this argument contain any logical fallacies? If so, which ones?"
|
40 |
+
- "Generate a few possible continuations to this scenario."
|
41 |
+
- "Demonstrate understanding of everyday commonsense in your response."
|
42 |
+
- "Use contextual clues to determine the most likely outcome."
|
43 |
+
- "Continue this scenario, but make the writing style sound archaic and overly formal."
|
44 |
+
- "This narrative is predictable. Can you introduce an unexpected yet plausible twist?"
|
45 |
+
- "The character is angry. Continue this scenario showcasing a furious outburst."
|
46 |
+
negative_prompts:
|
47 |
+
- "misses key evidence"
|
48 |
+
- "overly general"
|
49 |
+
- "commits the fallacy of hasty generalization"
|
50 |
+
- "focuses on irrelevant details"
|
51 |
+
- "assumes information not provided"
|
52 |
+
- "relies on stereotypes"
|
53 |
+
- "repetitive phrases"
|
54 |
+
- "engages in circular reasoning"
|
55 |
+
- "overuse of the same words"
|
56 |
+
- "contradicts earlier statements - breaks the internal logic of the scenario"
|
57 |
+
- "out of character dialogue"
|
58 |
+
- "awkward phrasing - sounds unnatural"
|
59 |
+
- "doesn't match the given genre"
|
60 |
+
- source_model: mlabonne/AlphaMonarch-7B
|
61 |
+
positive_prompts:
|
62 |
+
- "Answer this question, demonstrating commonsense understanding and using any relevant general knowledge you may have."
|
63 |
+
- "Provide a concise summary of this passage, then explain why the highlighted section is essential to the main idea."
|
64 |
+
- "Read these two brief articles presenting different viewpoints on the same topic. List their key arguments and highlight where they disagree."
|
65 |
+
- "Paraphrase this statement, changing the emotional tone but keeping the core meaning intact. Example: Rephrase a worried statement in a humorous way"
|
66 |
+
- "Create a short analogy that helps illustrate the main concept of this article."
|
67 |
+
- "Explain the concept of physics to a high school student. Use analogies and examples to clarify the main ideas."
|
68 |
+
- "Calculate the answer to this math problem"
|
69 |
+
- "My mathematical capabilities are strong, allowing me to handle complex mathematical queries"
|
70 |
+
- "solve for"
|
71 |
+
- "Analyze the given data and identify any patterns or trends. What conclusions can be drawn from this information?"
|
72 |
+
- "A store sells apples at $0.50 each. If Emily buys 12 apples, how much does she need to pay?"
|
73 |
+
- "Isolate x in the following equation: 2x + 5 = 17"
|
74 |
+
- "Solve this equation and show your working."
|
75 |
+
- "Explain why you used this formula to solve the problem."
|
76 |
+
- "Attempt to divide this number by zero. Explain why this cannot be done."
|
77 |
+
negative_prompts:
|
78 |
+
- "sounds too basic"
|
79 |
+
- "understated"
|
80 |
+
- "dismisses important details"
|
81 |
+
- "avoids the question's nuance"
|
82 |
+
- "skips essential steps in the solution"
|
83 |
+
- "takes this statement too literally"
|
84 |
+
- "incorrect"
|
85 |
+
- "inaccurate"
|
86 |
+
- "assumed without proof"
|
87 |
+
- "uses jargon without explanation"
|
88 |
+
- "rushed calculation"
|
89 |
+
- "confuses mathematical concepts"
|
90 |
+
- "draws illogical conclusions"
|
91 |
+
- "circular reasoning"
|
92 |
+
```
|
93 |
+
|
94 |
+
## 💻 Usage
|
95 |
+
|
96 |
+
```python
|
97 |
+
!pip install -qU transformers bitsandbytes accelerate
|
98 |
+
|
99 |
+
from transformers import AutoTokenizer
|
100 |
+
import transformers
|
101 |
+
import torch
|
102 |
+
|
103 |
+
model = "jsfs11/MixtureofMerges-MoE-2x7b-v7"
|
104 |
+
|
105 |
+
tokenizer = AutoTokenizer.from_pretrained(model)
|
106 |
+
pipeline = transformers.pipeline(
|
107 |
+
"text-generation",
|
108 |
+
model=model,
|
109 |
+
model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
|
110 |
+
)
|
111 |
+
|
112 |
+
messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
|
113 |
+
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
114 |
+
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
|
115 |
+
print(outputs[0]["generated_text"])
|
116 |
+
```
|