mlabonne leaderboard-pr-bot commited on
Commit
69b1b4a
1 Parent(s): 857e32a

Adding Evaluation Results (#10)

Browse files

- Adding Evaluation Results (4985d4e235a8166abba352e3b909c16884eabe32)


Co-authored-by: Open LLM Leaderboard PR Bot <leaderboard-pr-bot@users.noreply.huggingface.co>

Files changed (1) hide show
  1. README.md +119 -3
README.md CHANGED
@@ -1,7 +1,5 @@
1
  ---
2
  license: other
3
- license_name: microsoft-research-license
4
- license_link: https://huggingface.co/WizardLM/WizardMath-7B-V1.1/resolve/main/LICENSE
5
  tags:
6
  - moe
7
  - merge
@@ -11,6 +9,111 @@ tags:
11
  - beowolx/CodeNinja-1.0-OpenChat-7B
12
  - maywell/PiVoT-0.1-Starling-LM-RP
13
  - WizardLM/WizardMath-7B-V1.1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14
  ---
15
 
16
  ![](https://i.imgur.com/vq1QHEA.jpg)
@@ -181,4 +284,17 @@ print(outputs[0]["generated_text"])
181
 
182
  Output:
183
 
184
- > A Mixture of Experts (ME) is a machine learning technique that combines multiple expert models to make predictions or decisions. Each expert model is specialized in a different aspect of the problem, and their outputs are combined to produce a more accurate and robust solution. This approach allows the model to leverage the strengths of individual experts and compensate for their weaknesses, improving overall performance.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: other
 
 
3
  tags:
4
  - moe
5
  - merge
 
9
  - beowolx/CodeNinja-1.0-OpenChat-7B
10
  - maywell/PiVoT-0.1-Starling-LM-RP
11
  - WizardLM/WizardMath-7B-V1.1
12
+ license_name: microsoft-research-license
13
+ license_link: https://huggingface.co/WizardLM/WizardMath-7B-V1.1/resolve/main/LICENSE
14
+ model-index:
15
+ - name: Beyonder-4x7B-v2
16
+ results:
17
+ - task:
18
+ type: text-generation
19
+ name: Text Generation
20
+ dataset:
21
+ name: AI2 Reasoning Challenge (25-Shot)
22
+ type: ai2_arc
23
+ config: ARC-Challenge
24
+ split: test
25
+ args:
26
+ num_few_shot: 25
27
+ metrics:
28
+ - type: acc_norm
29
+ value: 68.77
30
+ name: normalized accuracy
31
+ source:
32
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Beyonder-4x7B-v2
33
+ name: Open LLM Leaderboard
34
+ - task:
35
+ type: text-generation
36
+ name: Text Generation
37
+ dataset:
38
+ name: HellaSwag (10-Shot)
39
+ type: hellaswag
40
+ split: validation
41
+ args:
42
+ num_few_shot: 10
43
+ metrics:
44
+ - type: acc_norm
45
+ value: 86.8
46
+ name: normalized accuracy
47
+ source:
48
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Beyonder-4x7B-v2
49
+ name: Open LLM Leaderboard
50
+ - task:
51
+ type: text-generation
52
+ name: Text Generation
53
+ dataset:
54
+ name: MMLU (5-Shot)
55
+ type: cais/mmlu
56
+ config: all
57
+ split: test
58
+ args:
59
+ num_few_shot: 5
60
+ metrics:
61
+ - type: acc
62
+ value: 65.1
63
+ name: accuracy
64
+ source:
65
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Beyonder-4x7B-v2
66
+ name: Open LLM Leaderboard
67
+ - task:
68
+ type: text-generation
69
+ name: Text Generation
70
+ dataset:
71
+ name: TruthfulQA (0-shot)
72
+ type: truthful_qa
73
+ config: multiple_choice
74
+ split: validation
75
+ args:
76
+ num_few_shot: 0
77
+ metrics:
78
+ - type: mc2
79
+ value: 60.68
80
+ source:
81
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Beyonder-4x7B-v2
82
+ name: Open LLM Leaderboard
83
+ - task:
84
+ type: text-generation
85
+ name: Text Generation
86
+ dataset:
87
+ name: Winogrande (5-shot)
88
+ type: winogrande
89
+ config: winogrande_xl
90
+ split: validation
91
+ args:
92
+ num_few_shot: 5
93
+ metrics:
94
+ - type: acc
95
+ value: 80.9
96
+ name: accuracy
97
+ source:
98
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Beyonder-4x7B-v2
99
+ name: Open LLM Leaderboard
100
+ - task:
101
+ type: text-generation
102
+ name: Text Generation
103
+ dataset:
104
+ name: GSM8k (5-shot)
105
+ type: gsm8k
106
+ config: main
107
+ split: test
108
+ args:
109
+ num_few_shot: 5
110
+ metrics:
111
+ - type: acc
112
+ value: 71.72
113
+ name: accuracy
114
+ source:
115
+ url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Beyonder-4x7B-v2
116
+ name: Open LLM Leaderboard
117
  ---
118
 
119
  ![](https://i.imgur.com/vq1QHEA.jpg)
 
284
 
285
  Output:
286
 
287
+ > A Mixture of Experts (ME) is a machine learning technique that combines multiple expert models to make predictions or decisions. Each expert model is specialized in a different aspect of the problem, and their outputs are combined to produce a more accurate and robust solution. This approach allows the model to leverage the strengths of individual experts and compensate for their weaknesses, improving overall performance.
288
+ # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
289
+ Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_mlabonne__Beyonder-4x7B-v2)
290
+
291
+ | Metric |Value|
292
+ |---------------------------------|----:|
293
+ |Avg. |72.33|
294
+ |AI2 Reasoning Challenge (25-Shot)|68.77|
295
+ |HellaSwag (10-Shot) |86.80|
296
+ |MMLU (5-Shot) |65.10|
297
+ |TruthfulQA (0-shot) |60.68|
298
+ |Winogrande (5-shot) |80.90|
299
+ |GSM8k (5-shot) |71.72|
300
+