Hugging Face's logo

saishf
/
Fimbulvetr-Kuro-Lotus-10.7B

Text Generation
Transformers
Safetensors
llama
mergekit
Merge
Eval Results
text-generation-inference
Inference Endpoints
Model card Files Community
4
saishf leaderboard-pr-bot commited on
Commit
2b0230e
•
1 Parent(s): 8ff080f

Adding Evaluation Results (#2)

Browse files

- Adding Evaluation Results (e1bc40a7cde92ff18fc658dc6786deea4066bef8)


Co-authored-by: Open LLM Leaderboard PR Bot <[email protected]>

Files changed (1) hide show
  1. README.md +121 -5
README.md CHANGED
@@ -1,12 +1,115 @@
1
  ---
2
- base_model:
3
- - Sao10K/Fimbulvetr-10.7B-v1
4
- - saishf/Kuro-Lotus-10.7B
5
  library_name: transformers
6
  tags:
7
  - mergekit
8
  - merge
9
- license: cc-by-nc-4.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10
  ---
11
  This model is a merge of my personal favourite models, i couldn't decide between them so why not have both? Without MOE cause gpu poor :3
12
 
@@ -69,4 +172,17 @@ parameters:
69
  - value: 0.5
70
  dtype: bfloat16
71
 
72
- ```
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: cc-by-nc-4.0
 
 
3
  library_name: transformers
4
  tags:
5
  - mergekit
6
  - merge
7
+ base_model:
8
+ - Sao10K/Fimbulvetr-10.7B-v1
9
+ - saishf/Kuro-Lotus-10.7B
10
+ model-index:
11
+ - name: Fimbulvetr-Kuro-Lotus-10.7B
12
+ results:
13
+ - task:
14
+ type: text-generation
15
+ name: Text Generation
16
+ dataset:
17
+ name: AI2 Reasoning Challenge (25-Shot)
18
+ type: ai2_arc
19
+ config: ARC-Challenge
20
+ split: test
21
+ args:
22
+ num_few_shot: 25
23
+ metrics:
24
+ - type: acc_norm
25
+ value: 69.54
26
+ name: normalized accuracy
27
+ source:
28
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
29
+ name: Open LLM Leaderboard
30
+ - task:
31
+ type: text-generation
32
+ name: Text Generation
33
+ dataset:
34
+ name: HellaSwag (10-Shot)
35
+ type: hellaswag
36
+ split: validation
37
+ args:
38
+ num_few_shot: 10
39
+ metrics:
40
+ - type: acc_norm
41
+ value: 87.87
42
+ name: normalized accuracy
43
+ source:
44
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
45
+ name: Open LLM Leaderboard
46
+ - task:
47
+ type: text-generation
48
+ name: Text Generation
49
+ dataset:
50
+ name: MMLU (5-Shot)
51
+ type: cais/mmlu
52
+ config: all
53
+ split: test
54
+ args:
55
+ num_few_shot: 5
56
+ metrics:
57
+ - type: acc
58
+ value: 66.99
59
+ name: accuracy
60
+ source:
61
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
62
+ name: Open LLM Leaderboard
63
+ - task:
64
+ type: text-generation
65
+ name: Text Generation
66
+ dataset:
67
+ name: TruthfulQA (0-shot)
68
+ type: truthful_qa
69
+ config: multiple_choice
70
+ split: validation
71
+ args:
72
+ num_few_shot: 0
73
+ metrics:
74
+ - type: mc2
75
+ value: 60.95
76
+ source:
77
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
78
+ name: Open LLM Leaderboard
79
+ - task:
80
+ type: text-generation
81
+ name: Text Generation
82
+ dataset:
83
+ name: Winogrande (5-shot)
84
+ type: winogrande
85
+ config: winogrande_xl
86
+ split: validation
87
+ args:
88
+ num_few_shot: 5
89
+ metrics:
90
+ - type: acc
91
+ value: 84.14
92
+ name: accuracy
93
+ source:
94
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
95
+ name: Open LLM Leaderboard
96
+ - task:
97
+ type: text-generation
98
+ name: Text Generation
99
+ dataset:
100
+ name: GSM8k (5-shot)
101
+ type: gsm8k
102
+ config: main
103
+ split: test
104
+ args:
105
+ num_few_shot: 5
106
+ metrics:
107
+ - type: acc
108
+ value: 66.87
109
+ name: accuracy
110
+ source:
111
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
112
+ name: Open LLM Leaderboard
113
  ---
114
  This model is a merge of my personal favourite models, i couldn't decide between them so why not have both? Without MOE cause gpu poor :3
115
 
 
172
  - value: 0.5
173
  dtype: bfloat16
174
 
175
+ ```
176
+ # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
177
+ Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_saishf__Fimbulvetr-Kuro-Lotus-10.7B)
178
+
179
+ | Metric |Value|
180
+ |---------------------------------|----:|
181
+ |Avg. |72.73|
182
+ |AI2 Reasoning Challenge (25-Shot)|69.54|
183
+ |HellaSwag (10-Shot) |87.87|
184
+ |MMLU (5-Shot) |66.99|
185
+ |TruthfulQA (0-shot) |60.95|
186
+ |Winogrande (5-shot) |84.14|
187
+ |GSM8k (5-shot) |66.87|
188
+