Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- README.md +83 -0
- beyonder-4x7b-v3.Q6_K.gguf +3 -0
.gitattributes
CHANGED
@@ -36,3 +36,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
36 |
beyonder-4x7b-v3.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
|
37 |
beyonder-4x7b-v3.Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
|
38 |
beyonder-4x7b-v3.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
|
|
|
36 |
beyonder-4x7b-v3.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
|
37 |
beyonder-4x7b-v3.Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
|
38 |
beyonder-4x7b-v3.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
39 |
+
beyonder-4x7b-v3.Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: cc-by-nc-4.0
|
3 |
+
tags:
|
4 |
+
- moe
|
5 |
+
- frankenmoe
|
6 |
+
- merge
|
7 |
+
- mergekit
|
8 |
+
- lazymergekit
|
9 |
+
- autoquant
|
10 |
+
- gguf
|
11 |
+
base_model:
|
12 |
+
- mlabonne/AlphaMonarch-7B
|
13 |
+
- beowolx/CodeNinja-1.0-OpenChat-7B
|
14 |
+
- SanjiWatsuki/Kunoichi-DPO-v2-7B
|
15 |
+
- mlabonne/NeuralDaredevil-7B
|
16 |
+
---
|
17 |
+
|
18 |
+
# Beyonder-4x7B-v3
|
19 |
+
|
20 |
+
Beyonder-4x7B-v3 is a Mixure of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
21 |
+
* [mlabonne/AlphaMonarch-7B](https://huggingface.co/mlabonne/AlphaMonarch-7B)
|
22 |
+
* [beowolx/CodeNinja-1.0-OpenChat-7B](https://huggingface.co/beowolx/CodeNinja-1.0-OpenChat-7B)
|
23 |
+
* [SanjiWatsuki/Kunoichi-DPO-v2-7B](https://huggingface.co/SanjiWatsuki/Kunoichi-DPO-v2-7B)
|
24 |
+
* [mlabonne/NeuralDaredevil-7B](https://huggingface.co/mlabonne/NeuralDaredevil-7B)
|
25 |
+
|
26 |
+
## 🧩 Configuration
|
27 |
+
|
28 |
+
```yaml
|
29 |
+
base_model: mlabonne/AlphaMonarch-7B
|
30 |
+
experts:
|
31 |
+
- source_model: mlabonne/AlphaMonarch-7B
|
32 |
+
positive_prompts:
|
33 |
+
- "chat"
|
34 |
+
- "assistant"
|
35 |
+
- "tell me"
|
36 |
+
- "explain"
|
37 |
+
- "I want"
|
38 |
+
- source_model: beowolx/CodeNinja-1.0-OpenChat-7B
|
39 |
+
positive_prompts:
|
40 |
+
- "code"
|
41 |
+
- "python"
|
42 |
+
- "javascript"
|
43 |
+
- "programming"
|
44 |
+
- "algorithm"
|
45 |
+
- source_model: SanjiWatsuki/Kunoichi-DPO-v2-7B
|
46 |
+
positive_prompts:
|
47 |
+
- "storywriting"
|
48 |
+
- "write"
|
49 |
+
- "scene"
|
50 |
+
- "story"
|
51 |
+
- "character"
|
52 |
+
- source_model: mlabonne/NeuralDaredevil-7B
|
53 |
+
positive_prompts:
|
54 |
+
- "reason"
|
55 |
+
- "math"
|
56 |
+
- "mathematics"
|
57 |
+
- "solve"
|
58 |
+
- "count"
|
59 |
+
```
|
60 |
+
|
61 |
+
## 💻 Usage
|
62 |
+
|
63 |
+
```python
|
64 |
+
!pip install -qU transformers bitsandbytes accelerate
|
65 |
+
|
66 |
+
from transformers import AutoTokenizer
|
67 |
+
import transformers
|
68 |
+
import torch
|
69 |
+
|
70 |
+
model = "mlabonne/Beyonder-4x7B-v3"
|
71 |
+
|
72 |
+
tokenizer = AutoTokenizer.from_pretrained(model)
|
73 |
+
pipeline = transformers.pipeline(
|
74 |
+
"text-generation",
|
75 |
+
model=model,
|
76 |
+
model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
|
77 |
+
)
|
78 |
+
|
79 |
+
messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
|
80 |
+
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
81 |
+
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
|
82 |
+
print(outputs[0]["generated_text"])
|
83 |
+
```
|
beyonder-4x7b-v3.Q6_K.gguf
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0d36dfe9cdfa82a0b6aeb6dc5f214fb9c98b7425433c9b58cfe2d62932fa3831
|
3 |
+
size 19815794656
|