Update README.md
Browse files
README.md
CHANGED
@@ -3,26 +3,70 @@ license: apache-2.0
|
|
3 |
tags:
|
4 |
- merge
|
5 |
- mergekit
|
6 |
-
-
|
7 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
- unsloth/Qwen2.5-Coder-1.5B-Instruct
|
9 |
-
- Qwen/Qwen2.5-Math-1.5B-Instruct
|
10 |
-
- bunnycore/Qwen2.5-1.5B-Matrix
|
11 |
-
- Syed-Hasan-8503/Qwen2.5-1.5B-Instruct-WO-Adam-mini
|
12 |
-
- Goekdeniz-Guelmez/Josiefied-Qwen2.5-1.5B-Instruct-abliterated-v3
|
13 |
---
|
14 |
|
15 |
-
# ZeroXClem/Qwen2.5-1.5B-Instruct-Coder-Math-Bunnycore-Fusion
|
16 |
|
17 |
-
ZeroXClem/Qwen2.5-1.5B-Instruct-Coder-Math-Bunnycore-Fusion
|
18 |
-
* [cyixiao/qwen-1.5B-openbookqa](https://huggingface.co/cyixiao/qwen-1.5B-openbookqa)
|
19 |
-
* [unsloth/Qwen2.5-Coder-1.5B-Instruct](https://huggingface.co/unsloth/Qwen2.5-Coder-1.5B-Instruct)
|
20 |
-
* [Qwen/Qwen2.5-Math-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B-Instruct)
|
21 |
-
* [bunnycore/Qwen2.5-1.5B-Matrix](https://huggingface.co/bunnycore/Qwen2.5-1.5B-Matrix)
|
22 |
-
* [Syed-Hasan-8503/Qwen2.5-1.5B-Instruct-WO-Adam-mini](https://huggingface.co/Syed-Hasan-8503/Qwen2.5-1.5B-Instruct-WO-Adam-mini)
|
23 |
-
* [Goekdeniz-Guelmez/Josiefied-Qwen2.5-1.5B-Instruct-abliterated-v3](https://huggingface.co/Goekdeniz-Guelmez/Josiefied-Qwen2.5-1.5B-Instruct-abliterated-v3)
|
24 |
|
25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
|
27 |
```yaml
|
28 |
merge_method: della
|
@@ -32,7 +76,7 @@ parameters:
|
|
32 |
lambda: 1.0
|
33 |
normalize: true
|
34 |
|
35 |
-
base_model: unsloth/Qwen2.5-Coder-1.5B-Instruct
|
36 |
|
37 |
models:
|
38 |
- model: cyixiao/qwen-1.5B-openbookqa
|
@@ -59,4 +103,45 @@ models:
|
|
59 |
parameters:
|
60 |
weight: 1
|
61 |
density: 0.5
|
62 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
tags:
|
4 |
- merge
|
5 |
- mergekit
|
6 |
+
- lazzymergekit
|
7 |
+
- Qwen2
|
8 |
+
- Coder
|
9 |
+
- Math
|
10 |
+
- Bunnycore
|
11 |
+
- Instruct
|
12 |
+
- OpenBookQA
|
13 |
+
- instruction-following
|
14 |
+
- long-form-generation
|
15 |
+
base_model:
|
16 |
- unsloth/Qwen2.5-Coder-1.5B-Instruct
|
|
|
|
|
|
|
|
|
17 |
---
|
18 |
|
|
|
19 |
|
20 |
+
# **ZeroXClem/Qwen2.5-1.5B-Instruct-Coder-Math-Bunnycore-Fusion**
|
|
|
|
|
|
|
|
|
|
|
|
|
21 |
|
22 |
+
**ZeroXClem/Qwen2.5-1.5B-Instruct-Coder-Math-Bunnycore-Fusion** is a cutting-edge merged model that combines the finest features from **instruction-following**, **coding**, **mathematical reasoning**, and **factual question-answering**. This powerhouse is designed for high performance in diverse technical, creative, and interactive tasks.
|
23 |
+
|
24 |
+
## 🌟 **Family Tree**
|
25 |
+
|
26 |
+
This model is the fusion of the following:
|
27 |
+
|
28 |
+
- [**cyixiao/qwen-1.5B-openbookqa**](https://huggingface.co/cyixiao/qwen-1.5B-openbookqa)
|
29 |
+
- [**unsloth/Qwen2.5-Coder-1.5B-Instruct**](https://huggingface.co/unsloth/Qwen2.5-Coder-1.5B-Instruct)
|
30 |
+
- [**Qwen/Qwen2.5-Math-1.5B-Instruct**](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B-Instruct)
|
31 |
+
- [**bunnycore/Qwen2.5-1.5B-Matrix**](https://huggingface.co/bunnycore/Qwen2.5-1.5B-Matrix)
|
32 |
+
- [**Syed-Hasan-8503/Qwen2.5-1.5B-Instruct-WO-Adam-mini**](https://huggingface.co/Syed-Hasan-8503/Qwen2.5-1.5B-Instruct-WO-Adam-mini)
|
33 |
+
- [**Goekdeniz-Guelmez/Josiefied-Qwen2.5-1.5B-Instruct-abliterated-v3**](https://huggingface.co/Goekdeniz-Guelmez/Josiefied-Qwen2.5-1.5B-Instruct-abliterated-v3)
|
34 |
+
|
35 |
+
These models have been seamlessly blended to create a versatile AI that excels across multiple domains.
|
36 |
+
|
37 |
+
---
|
38 |
+
|
39 |
+
## 🧬 **Detailed Model Lineage**
|
40 |
+
|
41 |
+
### **A: cyixiao/qwen-1.5B-openbookqa**
|
42 |
+
|
43 |
+
- Focuses on factual knowledge and reasoning from the OpenBookQA dataset, providing strong question-answering capabilities.
|
44 |
+
|
45 |
+
### **B: unsloth/Qwen2.5-Coder-1.5B-Instruct**
|
46 |
+
|
47 |
+
- Tailored for **coding** and **instruction-following**, this model enhances the ability to generate code and follow precise instructions with ease.
|
48 |
+
|
49 |
+
### **C: Qwen/Qwen2.5-Math-1.5B-Instruct**
|
50 |
+
|
51 |
+
- This model specializes in **mathematical reasoning** and logical problem-solving, making it perfect for structured tasks that require high-level thinking.
|
52 |
+
|
53 |
+
### **D: bunnycore/Qwen2.5-1.5B-Matrix**
|
54 |
+
|
55 |
+
- A multi-purpose model that blends **instruction**, **math**, and **coding**, providing a well-rounded performance in both structured and creative tasks.
|
56 |
+
|
57 |
+
### **E: Syed-Hasan-8503/Qwen2.5-1.5B-Instruct-WO-Adam-mini**
|
58 |
+
|
59 |
+
- Fine-tuned on conversational and identity-specific tasks, this model contributes to the model’s ability to handle **conversation-heavy** tasks with clarity.
|
60 |
+
|
61 |
+
### **F: Goekdeniz-Guelmez/Josiefied-Qwen2.5-1.5B-Instruct-abliterated-v3**
|
62 |
+
|
63 |
+
- This model brings **uncensored** capabilities, ensuring that the AI is flexible and adaptable in open-ended and unrestricted instruction-following scenarios.
|
64 |
+
|
65 |
+
---
|
66 |
+
|
67 |
+
## 🛠️ **Merge Details**
|
68 |
+
|
69 |
+
The model was merged using the **DELLA merge method** with **bfloat16** precision, ensuring high-performance across multiple task types. Here's the configuration used for the merge:
|
70 |
|
71 |
```yaml
|
72 |
merge_method: della
|
|
|
76 |
lambda: 1.0
|
77 |
normalize: true
|
78 |
|
79 |
+
base_model: unsloth/Qwen2.5-Coder-1.5B-Instruct
|
80 |
|
81 |
models:
|
82 |
- model: cyixiao/qwen-1.5B-openbookqa
|
|
|
103 |
parameters:
|
104 |
weight: 1
|
105 |
density: 0.5
|
106 |
+
|
107 |
+
```
|
108 |
+
|
109 |
+
---
|
110 |
+
|
111 |
+
## 🎯 **Key Features & Capabilities**
|
112 |
+
|
113 |
+
### **1. Coding and Instruction Following**:
|
114 |
+
|
115 |
+
This model excels in technical coding tasks thanks to the contributions from **Qwen2.5-Coder** and **Matrix**.
|
116 |
+
|
117 |
+
### **2. Mathematical Reasoning**:
|
118 |
+
|
119 |
+
With **Qwen2.5-Math-1.5B-Instruct**, the model is perfect for solving complex **mathematical problems** and structured logical tasks.
|
120 |
+
|
121 |
+
### **3. Conversational Abilities**:
|
122 |
+
|
123 |
+
Fine-tuned on conversation and identity tasks, the model handles complex dialogue and conversational exchanges with **Syed-Hasan-8503**.
|
124 |
+
|
125 |
+
### **4. Uncensored Versatility**:
|
126 |
+
|
127 |
+
Thanks to **Josiefied-Qwen2.5**, this model can operate without restrictions, making it ideal for **open-ended instruction-following**.
|
128 |
+
|
129 |
+
---
|
130 |
+
|
131 |
+
## 📜 **License**
|
132 |
+
|
133 |
+
This model is open-sourced under the **Apache-2.0 License**, allowing others to use and modify it freely, as long as they give proper attribution.
|
134 |
+
|
135 |
+
---
|
136 |
+
|
137 |
+
## 💡 **Tags**
|
138 |
+
|
139 |
+
- `merge`
|
140 |
+
- `Qwen`
|
141 |
+
- `Coder`
|
142 |
+
- `Math`
|
143 |
+
- `Bunnycore`
|
144 |
+
- `instruction-following`
|
145 |
+
- `long-form-generation`
|
146 |
+
|
147 |
+
---
|