Text Generation
Transformers
Safetensors
mixtral
Mixture of Experts
frankenmoe
Merge
mergekit
lazymergekit
Locutusque/TinyMistral-248M-v2
Locutusque/TinyMistral-248M-v2.5
Locutusque/TinyMistral-248M-v2.5-Instruct
jtatman/tinymistral-v2-pycoder-instruct-248m
Felladrin/TinyMistral-248M-SFT-v4
Locutusque/TinyMistral-248M-v2-Instruct
text-generation-inference
Inference Endpoints
Locutusque
commited on
Commit
•
9aa780b
1
Parent(s):
1a1b29f
Upload folder using huggingface_hub
Browse files- README.md +59 -43
- mergekit_moe_config.yml +58 -30
- model-00001-of-00001.safetensors +1 -1
README.md
CHANGED
@@ -19,23 +19,11 @@ base_model:
|
|
19 |
- jtatman/tinymistral-v2-pycoder-instruct-248m
|
20 |
- Felladrin/TinyMistral-248M-SFT-v4
|
21 |
- Locutusque/TinyMistral-248M-v2-Instruct
|
22 |
-
language:
|
23 |
-
- en
|
24 |
-
inference:
|
25 |
-
parameters:
|
26 |
-
do_sample: true
|
27 |
-
temperature: 0.2
|
28 |
-
top_p: 0.14
|
29 |
-
top_k: 12
|
30 |
-
max_new_tokens: 250
|
31 |
-
repetition_penalty: 1.15
|
32 |
-
widget:
|
33 |
-
- text: "<|im_start|>user\nWrite me a Python program that calculates the factorial of n. <|im_end|>\n<|im_start|>assistant\n"
|
34 |
---
|
35 |
|
36 |
# TinyMistral-6x248M
|
37 |
|
38 |
-
TinyMistral-6x248M is a Mixure of Experts (MoE)
|
39 |
* [Locutusque/TinyMistral-248M-v2](https://huggingface.co/Locutusque/TinyMistral-248M-v2)
|
40 |
* [Locutusque/TinyMistral-248M-v2.5](https://huggingface.co/Locutusque/TinyMistral-248M-v2.5)
|
41 |
* [Locutusque/TinyMistral-248M-v2.5-Instruct](https://huggingface.co/Locutusque/TinyMistral-248M-v2.5-Instruct)
|
@@ -50,46 +38,74 @@ base_model: Locutusque/TinyMistral-248M-v2.5
|
|
50 |
experts:
|
51 |
- source_model: Locutusque/TinyMistral-248M-v2
|
52 |
positive_prompts:
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
|
|
|
|
|
|
|
|
57 |
- source_model: Locutusque/TinyMistral-248M-v2.5
|
58 |
positive_prompts:
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
- source_model: Locutusque/TinyMistral-248M-v2.5-Instruct
|
64 |
positive_prompts:
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
|
|
|
|
|
|
|
|
71 |
- source_model: jtatman/tinymistral-v2-pycoder-instruct-248m
|
72 |
positive_prompts:
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
|
|
|
|
|
|
|
|
|
|
|
77 |
- source_model: Felladrin/TinyMistral-248M-SFT-v4
|
78 |
positive_prompts:
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
|
84 |
-
|
|
|
|
|
|
|
|
|
|
|
85 |
- source_model: Locutusque/TinyMistral-248M-v2-Instruct
|
86 |
positive_prompts:
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
|
|
|
|
|
|
93 |
```
|
94 |
|
95 |
## 💻 Usage
|
|
|
19 |
- jtatman/tinymistral-v2-pycoder-instruct-248m
|
20 |
- Felladrin/TinyMistral-248M-SFT-v4
|
21 |
- Locutusque/TinyMistral-248M-v2-Instruct
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
---
|
23 |
|
24 |
# TinyMistral-6x248M
|
25 |
|
26 |
+
TinyMistral-6x248M is a Mixure of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
27 |
* [Locutusque/TinyMistral-248M-v2](https://huggingface.co/Locutusque/TinyMistral-248M-v2)
|
28 |
* [Locutusque/TinyMistral-248M-v2.5](https://huggingface.co/Locutusque/TinyMistral-248M-v2.5)
|
29 |
* [Locutusque/TinyMistral-248M-v2.5-Instruct](https://huggingface.co/Locutusque/TinyMistral-248M-v2.5-Instruct)
|
|
|
38 |
experts:
|
39 |
- source_model: Locutusque/TinyMistral-248M-v2
|
40 |
positive_prompts:
|
41 |
+
- "An emerging trend in global economics is"
|
42 |
+
- "TITLE: The Next Generation of Internet Connectivity"
|
43 |
+
- "begin a comprehensive analysis on the sociopolitical effects of"
|
44 |
+
negative_prompts:
|
45 |
+
- "Code a simple"
|
46 |
+
- "Explain the Krebs cycle in detail"
|
47 |
+
- "Compose a sonnet about"
|
48 |
+
|
49 |
- source_model: Locutusque/TinyMistral-248M-v2.5
|
50 |
positive_prompts:
|
51 |
+
- "Advanced C++ memory management techniques"
|
52 |
+
- "C# asynchronous programming best practices"
|
53 |
+
- "AI's role in predictive analytics"
|
54 |
+
- "textbook review on machine learning algorithms"
|
55 |
+
- "## Exercise: Design a C# interface for a CRM system"
|
56 |
+
- "## Solution: Optimize an AI-powered recommendation engine"
|
57 |
+
negative_prompts:
|
58 |
+
- "Narrate the story of"
|
59 |
+
- "The ethical considerations in"
|
60 |
+
- "Review the latest art exhibition by"
|
61 |
+
|
62 |
- source_model: Locutusque/TinyMistral-248M-v2.5-Instruct
|
63 |
positive_prompts:
|
64 |
+
- "What is the chemical formula for photosynthesis?"
|
65 |
+
- "Identification of a new mineral found on Mars"
|
66 |
+
- "physics: Explaining the concept of relativity"
|
67 |
+
- "Solve for x using differential equations:"
|
68 |
+
- "history: Analyze the causes of the French Revolution"
|
69 |
+
negative_prompts:
|
70 |
+
- "Devise a business plan for"
|
71 |
+
- "The evolution of culinary arts"
|
72 |
+
- "Orchestrate a piece for a string quartet"
|
73 |
+
|
74 |
- source_model: jtatman/tinymistral-v2-pycoder-instruct-248m
|
75 |
positive_prompts:
|
76 |
+
- "Write a Python program for facial recognition"
|
77 |
+
- "Explain dynamic typing in programming languages"
|
78 |
+
- "algorithm development for efficient data sorting"
|
79 |
+
negative_prompts:
|
80 |
+
- "Who was the first Emperor of Rome?"
|
81 |
+
- "Discuss the political dynamics in"
|
82 |
+
- "Provide a proof for Fermat's Last Theorem"
|
83 |
+
- "physics: The principles of thermodynamics"
|
84 |
+
|
85 |
- source_model: Felladrin/TinyMistral-248M-SFT-v4
|
86 |
positive_prompts:
|
87 |
+
- "Escreba sobre a influência da música no Brasil"
|
88 |
+
- "Voici un guide pour les voyageurs en France"
|
89 |
+
- "Para entender la política de México, se debe considerar"
|
90 |
+
- "Cuales son los efectos de la globalización en Argentina"
|
91 |
+
- "Welche gesellschaftlichen Veränderungen gibt es in Deutschland"
|
92 |
+
- "If you had to imagine a utopian city, what would be its core values?"
|
93 |
+
negative_prompts:
|
94 |
+
- "Calculate the integral of"
|
95 |
+
- "Describe the process of cell division"
|
96 |
+
- "Review the latest advancements in quantum computing"
|
97 |
+
|
98 |
- source_model: Locutusque/TinyMistral-248M-v2-Instruct
|
99 |
positive_prompts:
|
100 |
+
- "Write an essay on the evolution of international trade laws"
|
101 |
+
- "What are the key components of a sustainable urban ecosystem?"
|
102 |
+
- "instruct on effective negotiation techniques in diplomacy"
|
103 |
+
- "How does cognitive bias affect decision making in high-pressure environments?"
|
104 |
+
- "Identify the architectural significance of the Sydney Opera House"
|
105 |
+
negative_prompts:
|
106 |
+
- "Develop a script to automate"
|
107 |
+
- "Understanding inheritance in object-oriented programming"
|
108 |
+
- "philosophy of existentialism in contemporary society"
|
109 |
```
|
110 |
|
111 |
## 💻 Usage
|
mergekit_moe_config.yml
CHANGED
@@ -3,43 +3,71 @@ base_model: Locutusque/TinyMistral-248M-v2.5
|
|
3 |
experts:
|
4 |
- source_model: Locutusque/TinyMistral-248M-v2
|
5 |
positive_prompts:
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
|
|
|
|
|
|
|
|
10 |
- source_model: Locutusque/TinyMistral-248M-v2.5
|
11 |
positive_prompts:
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
- source_model: Locutusque/TinyMistral-248M-v2.5-Instruct
|
17 |
positive_prompts:
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
|
|
|
|
|
|
|
|
24 |
- source_model: jtatman/tinymistral-v2-pycoder-instruct-248m
|
25 |
positive_prompts:
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
|
|
|
|
|
|
|
|
|
|
30 |
- source_model: Felladrin/TinyMistral-248M-SFT-v4
|
31 |
positive_prompts:
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
|
|
|
|
|
|
|
|
|
|
38 |
- source_model: Locutusque/TinyMistral-248M-v2-Instruct
|
39 |
positive_prompts:
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
|
|
|
|
|
|
|
3 |
experts:
|
4 |
- source_model: Locutusque/TinyMistral-248M-v2
|
5 |
positive_prompts:
|
6 |
+
- "An emerging trend in global economics is"
|
7 |
+
- "TITLE: The Next Generation of Internet Connectivity"
|
8 |
+
- "begin a comprehensive analysis on the sociopolitical effects of"
|
9 |
+
negative_prompts:
|
10 |
+
- "Code a simple"
|
11 |
+
- "Explain the Krebs cycle in detail"
|
12 |
+
- "Compose a sonnet about"
|
13 |
+
|
14 |
- source_model: Locutusque/TinyMistral-248M-v2.5
|
15 |
positive_prompts:
|
16 |
+
- "Advanced C++ memory management techniques"
|
17 |
+
- "C# asynchronous programming best practices"
|
18 |
+
- "AI's role in predictive analytics"
|
19 |
+
- "textbook review on machine learning algorithms"
|
20 |
+
- "## Exercise: Design a C# interface for a CRM system"
|
21 |
+
- "## Solution: Optimize an AI-powered recommendation engine"
|
22 |
+
negative_prompts:
|
23 |
+
- "Narrate the story of"
|
24 |
+
- "The ethical considerations in"
|
25 |
+
- "Review the latest art exhibition by"
|
26 |
+
|
27 |
- source_model: Locutusque/TinyMistral-248M-v2.5-Instruct
|
28 |
positive_prompts:
|
29 |
+
- "What is the chemical formula for photosynthesis?"
|
30 |
+
- "Identification of a new mineral found on Mars"
|
31 |
+
- "physics: Explaining the concept of relativity"
|
32 |
+
- "Solve for x using differential equations:"
|
33 |
+
- "history: Analyze the causes of the French Revolution"
|
34 |
+
negative_prompts:
|
35 |
+
- "Devise a business plan for"
|
36 |
+
- "The evolution of culinary arts"
|
37 |
+
- "Orchestrate a piece for a string quartet"
|
38 |
+
|
39 |
- source_model: jtatman/tinymistral-v2-pycoder-instruct-248m
|
40 |
positive_prompts:
|
41 |
+
- "Write a Python program for facial recognition"
|
42 |
+
- "Explain dynamic typing in programming languages"
|
43 |
+
- "algorithm development for efficient data sorting"
|
44 |
+
negative_prompts:
|
45 |
+
- "Who was the first Emperor of Rome?"
|
46 |
+
- "Discuss the political dynamics in"
|
47 |
+
- "Provide a proof for Fermat's Last Theorem"
|
48 |
+
- "physics: The principles of thermodynamics"
|
49 |
+
|
50 |
- source_model: Felladrin/TinyMistral-248M-SFT-v4
|
51 |
positive_prompts:
|
52 |
+
- "Escreba sobre a influência da música no Brasil"
|
53 |
+
- "Voici un guide pour les voyageurs en France"
|
54 |
+
- "Para entender la política de México, se debe considerar"
|
55 |
+
- "Cuales son los efectos de la globalización en Argentina"
|
56 |
+
- "Welche gesellschaftlichen Veränderungen gibt es in Deutschland"
|
57 |
+
- "If you had to imagine a utopian city, what would be its core values?"
|
58 |
+
negative_prompts:
|
59 |
+
- "Calculate the integral of"
|
60 |
+
- "Describe the process of cell division"
|
61 |
+
- "Review the latest advancements in quantum computing"
|
62 |
+
|
63 |
- source_model: Locutusque/TinyMistral-248M-v2-Instruct
|
64 |
positive_prompts:
|
65 |
+
- "Write an essay on the evolution of international trade laws"
|
66 |
+
- "What are the key components of a sustainable urban ecosystem?"
|
67 |
+
- "instruct on effective negotiation techniques in diplomacy"
|
68 |
+
- "How does cognitive bias affect decision making in high-pressure environments?"
|
69 |
+
- "Identify the architectural significance of the Sydney Opera House"
|
70 |
+
negative_prompts:
|
71 |
+
- "Develop a script to automate"
|
72 |
+
- "Understanding inheritance in object-oriented programming"
|
73 |
+
- "philosophy of existentialism in contemporary society"
|
model-00001-of-00001.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 2006182688
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9ffeb76d03362dac81d289a80a9d0f32b2d6d62bed247cfbb61137d59bc2ae4f
|
3 |
size 2006182688
|