Update README.md
Browse files
README.md
CHANGED
@@ -1,15 +1,6 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
library_name: transformers
|
4 |
-
tags:
|
5 |
-
- merge
|
6 |
-
- mergekit
|
7 |
-
- lazymergekit
|
8 |
-
- lemon07r/Gemma-2-Ataraxy-9B
|
9 |
-
- TheDrummer/Gemmasutra-9B-v1
|
10 |
-
base_model:
|
11 |
-
- lemon07r/Gemma-2-Ataraxy-9B
|
12 |
-
- TheDrummer/Gemmasutra-9B-v1
|
13 |
model-index:
|
14 |
- name: Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
15 |
results:
|
@@ -26,7 +17,8 @@ model-index:
|
|
26 |
value: 76.49
|
27 |
name: strict accuracy
|
28 |
source:
|
29 |
-
url:
|
|
|
30 |
name: Open LLM Leaderboard
|
31 |
- task:
|
32 |
type: text-generation
|
@@ -41,7 +33,8 @@ model-index:
|
|
41 |
value: 42.25
|
42 |
name: normalized accuracy
|
43 |
source:
|
44 |
-
url:
|
|
|
45 |
name: Open LLM Leaderboard
|
46 |
- task:
|
47 |
type: text-generation
|
@@ -56,7 +49,8 @@ model-index:
|
|
56 |
value: 1.74
|
57 |
name: exact match
|
58 |
source:
|
59 |
-
url:
|
|
|
60 |
name: Open LLM Leaderboard
|
61 |
- task:
|
62 |
type: text-generation
|
@@ -71,7 +65,8 @@ model-index:
|
|
71 |
value: 10.74
|
72 |
name: acc_norm
|
73 |
source:
|
74 |
-
url:
|
|
|
75 |
name: Open LLM Leaderboard
|
76 |
- task:
|
77 |
type: text-generation
|
@@ -86,7 +81,8 @@ model-index:
|
|
86 |
value: 12.39
|
87 |
name: acc_norm
|
88 |
source:
|
89 |
-
url:
|
|
|
90 |
name: Open LLM Leaderboard
|
91 |
- task:
|
92 |
type: text-generation
|
@@ -103,36 +99,13 @@ model-index:
|
|
103 |
value: 35.63
|
104 |
name: accuracy
|
105 |
source:
|
106 |
-
url:
|
|
|
107 |
name: Open LLM Leaderboard
|
108 |
---
|
109 |
|
110 |
# Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
111 |
|
112 |
-
Gemma-2-Ataraxy-Gemmasutra-9B-slerp is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
113 |
-
* [lemon07r/Gemma-2-Ataraxy-9B](https://huggingface.co/lemon07r/Gemma-2-Ataraxy-9B)
|
114 |
-
* [TheDrummer/Gemmasutra-9B-v1](https://huggingface.co/TheDrummer/Gemmasutra-9B-v1)
|
115 |
-
|
116 |
-
## 🧩 Configuration
|
117 |
-
|
118 |
-
```yaml
|
119 |
-
slices:
|
120 |
-
- sources:
|
121 |
-
- model: lemon07r/Gemma-2-Ataraxy-9B
|
122 |
-
layer_range: [0, 42]
|
123 |
-
- model: TheDrummer/Gemmasutra-9B-v1
|
124 |
-
layer_range: [0, 42]
|
125 |
-
merge_method: slerp
|
126 |
-
base_model: lemon07r/Gemma-2-Ataraxy-9B
|
127 |
-
parameters:
|
128 |
-
t:
|
129 |
-
- filter: self_attn
|
130 |
-
value: [0, 0.5, 0.3, 0.7, 1]
|
131 |
-
- filter: mlp
|
132 |
-
value: [1, 0.5, 0.7, 0.3, 0]
|
133 |
-
- value: 0.5
|
134 |
-
dtype: bfloat16
|
135 |
-
```
|
136 |
|
137 |
## 💻 Usage
|
138 |
|
@@ -169,5 +142,4 @@ Detailed results can be found [here](https://huggingface.co/datasets/open-llm-le
|
|
169 |
|MATH Lvl 5 (4-Shot)| 1.74|
|
170 |
|GPQA (0-shot) |10.74|
|
171 |
|MuSR (0-shot) |12.39|
|
172 |
-
|MMLU-PRO (5-shot) |35.63|
|
173 |
-
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
library_name: transformers
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
model-index:
|
5 |
- name: Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
6 |
results:
|
|
|
17 |
value: 76.49
|
18 |
name: strict accuracy
|
19 |
source:
|
20 |
+
url: >-
|
21 |
+
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
22 |
name: Open LLM Leaderboard
|
23 |
- task:
|
24 |
type: text-generation
|
|
|
33 |
value: 42.25
|
34 |
name: normalized accuracy
|
35 |
source:
|
36 |
+
url: >-
|
37 |
+
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
38 |
name: Open LLM Leaderboard
|
39 |
- task:
|
40 |
type: text-generation
|
|
|
49 |
value: 1.74
|
50 |
name: exact match
|
51 |
source:
|
52 |
+
url: >-
|
53 |
+
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
54 |
name: Open LLM Leaderboard
|
55 |
- task:
|
56 |
type: text-generation
|
|
|
65 |
value: 10.74
|
66 |
name: acc_norm
|
67 |
source:
|
68 |
+
url: >-
|
69 |
+
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
70 |
name: Open LLM Leaderboard
|
71 |
- task:
|
72 |
type: text-generation
|
|
|
81 |
value: 12.39
|
82 |
name: acc_norm
|
83 |
source:
|
84 |
+
url: >-
|
85 |
+
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
86 |
name: Open LLM Leaderboard
|
87 |
- task:
|
88 |
type: text-generation
|
|
|
99 |
value: 35.63
|
100 |
name: accuracy
|
101 |
source:
|
102 |
+
url: >-
|
103 |
+
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
104 |
name: Open LLM Leaderboard
|
105 |
---
|
106 |
|
107 |
# Gemma-2-Ataraxy-Gemmasutra-9B-slerp
|
108 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
109 |
|
110 |
## 💻 Usage
|
111 |
|
|
|
142 |
|MATH Lvl 5 (4-Shot)| 1.74|
|
143 |
|GPQA (0-shot) |10.74|
|
144 |
|MuSR (0-shot) |12.39|
|
145 |
+
|MMLU-PRO (5-shot) |35.63|
|
|