ZeroXClem commited on
Commit
24ea6c2
β€’
1 Parent(s): e81816e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +122 -4
README.md CHANGED
@@ -4,16 +4,53 @@ tags:
4
  - merge
5
  - mergekit
6
  - lazymergekit
 
 
 
 
 
 
 
 
 
 
 
7
  ---
8
 
9
  # ZeroXClem/L3.1-Aspire-Heart-Matrix-8B
10
 
11
- ZeroXClem/L3.1-Aspire-Heart-Matrix-8B is a merge of the following models using [mergekit](https://github.com/cg123/mergekit):
12
 
13
- ## 🧩 Configuration
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14
 
15
  ```yaml
16
- # Merge configuration for ZeroXClem/L3.1-Aspire-Heart-Matrix-8B using Model Stock
17
  models:
18
  - model: DreadPoor/Aspire-8B-model_stock
19
  - model: DreadPoor/Heart_Stolen-8B-Model_Stock
@@ -23,6 +60,87 @@ base_model: Khetterman/CursedMatrix-8B-v9
23
  normalize: false
24
  int8_mask: true
25
  dtype: bfloat16
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
26
 
 
27
 
28
- ```
 
4
  - merge
5
  - mergekit
6
  - lazymergekit
7
+ - vllm
8
+ - bfloat16
9
+ - llama
10
+ language:
11
+ - en
12
+ base_model:
13
+ - DreadPoor/Aspire-8B-model_stock
14
+ - DreadPoor/Heart_Stolen-8B-Model_Stock
15
+ - Khetterman/CursedMatrix-8B-v9
16
+ pipeline_tag: text-generation
17
+ library_name: transformers
18
  ---
19
 
20
  # ZeroXClem/L3.1-Aspire-Heart-Matrix-8B
21
 
22
+ **ZeroXClem/L3.1-Aspire-Heart-Matrix-8B** is an experimental language model crafted by merging three high-quality 8B parameter models using the **Model Stock Merge** method. This synthesis leverages the unique strengths of Aspire, Heart Stolen, and CursedMatrix, creating a highly versatile and robust language model for a wide array of tasks.
23
 
24
+
25
+ ## 🌟 Model Details
26
+
27
+ - **Name:** `ZeroXClem/L3.1-Aspire-Heart-Matrix-8B`
28
+ - **Base Model:** `Khetterman/CursedMatrix-8B-v9`
29
+ - **Merge Method:** `Model Stock`
30
+ - **Parameter Count:** `8 billion`
31
+ - **Precision:** `bfloat16`
32
+
33
+ ---
34
+
35
+ ## πŸ“‹ Models Used in the Merge
36
+
37
+ 1. **[Aspire](https://huggingface.co/DreadPoor/Aspire-8B-model_stock)**
38
+ *Creator: DreadPoor*
39
+ Known for exceptional performance across diverse tasks and benchmarks.
40
+
41
+ 2. **[Heart Stolen](https://huggingface.co/DreadPoor/Heart_Stolen-8B-Model_Stock)**
42
+ *Creator: DreadPoor*
43
+ Renowned for its creative and empathetic prowess.
44
+
45
+ 3. **[CursedMatrix](https://huggingface.co/Khetterman/CursedMatrix-8B-v9)**
46
+ *Creator: Khetterman*
47
+ Famous for its depth and complexity, particularly in creative writing and roleplay.
48
+
49
+ ---
50
+
51
+ ## βš™οΈ Merge Configuration
52
 
53
  ```yaml
 
54
  models:
55
  - model: DreadPoor/Aspire-8B-model_stock
56
  - model: DreadPoor/Heart_Stolen-8B-Model_Stock
 
60
  normalize: false
61
  int8_mask: true
62
  dtype: bfloat16
63
+ ```
64
+
65
+ ---
66
+
67
+ ## 🌌 Model Capabilities
68
+
69
+ This powerful merger unites the best features of its components:
70
+
71
+ - **Aspire**: Outstanding performance across general tasks and benchmarks.
72
+ - **Heart Stolen**: Creativity and empathy at its core.
73
+ - **CursedMatrix**: Mastery of complex and dynamic text generation.
74
+
75
+ The resulting model excels in:
76
+
77
+ - 🌟 **General Question Answering**
78
+ - πŸ“ **Creative Writing**
79
+ - βœ‚οΈ **Summarizing Long-Form Content**
80
+ - 🎭 **Roleplay Scenarios**
81
+ - βœ… **Task Completion and Problem-Solving**
82
+
83
+ ---
84
+
85
+ ## πŸ› οΈ Usage
86
+
87
+ This model is compatible with popular inference frameworks, including:
88
+
89
+ - [vLLM](https://github.com/vllm-project/vllm)
90
+ - [LMStudio](https://lmstudio.ai/)
91
+ - Hugging Face Transformers and other major libraries.
92
+
93
+
94
+ ```python
95
+ from transformers import AutoTokenizer, AutoModelForCausalLM
96
+
97
+ model_name = "ZeroXClem/Qwen2.5-7B-Qandora-CySec"
98
+ tokenizer = AutoTokenizer.from_pretrained(model_name)
99
+ model = AutoModelForCausalLM.from_pretrained(model_name)
100
+
101
+ input_text = "What are the fundamentals of python programming?"
102
+ input_ids = tokenizer.encode(input_text, return_tensors="pt")
103
+ output = model.generate(input_ids, max_length=100)
104
+ response = tokenizer.decode(output[0], skip_special_tokens=True)
105
+ print(response)
106
+ ```
107
+
108
+ Whether you're fine-tuning for specific tasks or using it out of the box, this model is a good base for your applications.
109
+
110
+ *Please give us any feedback if issues arise during inference via the discussions tab.*
111
+
112
+ ---
113
+
114
+ ## βš–οΈ Ethical Considerations
115
+
116
+ Given its uncensored origins and the potential for emergent behaviors, users should exercise caution. Be mindful of:
117
+
118
+ - Potential biases in outputs.
119
+ - Unexpected or unpredictable behavior in uncensored settings.
120
+
121
+ **Best Practices:** Implement robust content filtering and ensure responsible deployment in production environments.
122
+
123
+ ---
124
+
125
+ ## πŸ™ Acknowledgements
126
+
127
+ A heartfelt thank-you to the creators of the original models:
128
+
129
+ - **DreadPoor** for [Aspire](https://huggingface.co/DreadPoor/Aspire-8B-model_stock) and [Heart Stolen](https://huggingface.co/DreadPoor/Heart_Stolen-8B-Model_Stock).
130
+ - **Khetterman** for [CursedMatrix](https://huggingface.co/Khetterman/CursedMatrix-8B-v9).
131
+
132
+ Your brilliant contributions made this merge a reality.
133
+
134
+ ---
135
+
136
+ ## πŸ“œ License
137
+
138
+ This model inherits the licensing terms of its base components. Please refer to the licenses of:
139
+
140
+ - [Aspire](https://huggingface.co/DreadPoor/Aspire-8B-model_stock)
141
+ - [Heart Stolen](https://huggingface.co/DreadPoor/Heart_Stolen-8B-Model_Stock)
142
+ - [CursedMatrix](https://huggingface.co/Khetterman/CursedMatrix-8B-v9)
143
 
144
+ Ensure compliance with all licensing requirements when using this model.
145
 
146
+ ---