Hugging Face's logo

theprint
/
ReWiz-Qwen-2.5-14B

Text Generation
Transformers
Safetensors
GGUF
English
qwen2
text-generation-inference
unsloth
trl
sft
theprint
rewiz
Eval Results
Inference Endpoints
4-bit precision
bitsandbytes
Model card Files Community
1
theprint leaderboard-pr-bot commited on
Commit
ead62e8
•
1 Parent(s): bcc1701

Adding Evaluation Results (#1)

Browse files

- Adding Evaluation Results (a0c4e65ee25a9b0641bcf2fcd4d51da412556862)


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

Files changed (1) hide show
  1. README.md +110 -2
README.md CHANGED
@@ -1,5 +1,4 @@
1
  ---
2
- base_model: unsloth/qwen2.5-14b-bnb-4bit
3
  language:
4
  - en
5
  license: apache-2.0
@@ -12,8 +11,104 @@ tags:
12
  - sft
13
  - theprint
14
  - rewiz
 
15
  datasets:
16
  - theprint/ReWiz
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
17
  ---
18
 
19
  # Uploaded model
@@ -24,4 +119,17 @@ datasets:
24
 
25
  This qwen2 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
26
 
27
- [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
 
2
  language:
3
  - en
4
  license: apache-2.0
 
11
  - sft
12
  - theprint
13
  - rewiz
14
+ base_model: unsloth/qwen2.5-14b-bnb-4bit
15
  datasets:
16
  - theprint/ReWiz
17
+ model-index:
18
+ - name: ReWiz-Qwen-2.5-14B
19
+ results:
20
+ - task:
21
+ type: text-generation
22
+ name: Text Generation
23
+ dataset:
24
+ name: IFEval (0-Shot)
25
+ type: HuggingFaceH4/ifeval
26
+ args:
27
+ num_few_shot: 0
28
+ metrics:
29
+ - type: inst_level_strict_acc and prompt_level_strict_acc
30
+ value: 27.85
31
+ name: strict accuracy
32
+ source:
33
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-Qwen-2.5-14B
34
+ name: Open LLM Leaderboard
35
+ - task:
36
+ type: text-generation
37
+ name: Text Generation
38
+ dataset:
39
+ name: BBH (3-Shot)
40
+ type: BBH
41
+ args:
42
+ num_few_shot: 3
43
+ metrics:
44
+ - type: acc_norm
45
+ value: 44.86
46
+ name: normalized accuracy
47
+ source:
48
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-Qwen-2.5-14B
49
+ name: Open LLM Leaderboard
50
+ - task:
51
+ type: text-generation
52
+ name: Text Generation
53
+ dataset:
54
+ name: MATH Lvl 5 (4-Shot)
55
+ type: hendrycks/competition_math
56
+ args:
57
+ num_few_shot: 4
58
+ metrics:
59
+ - type: exact_match
60
+ value: 26.89
61
+ name: exact match
62
+ source:
63
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-Qwen-2.5-14B
64
+ name: Open LLM Leaderboard
65
+ - task:
66
+ type: text-generation
67
+ name: Text Generation
68
+ dataset:
69
+ name: GPQA (0-shot)
70
+ type: Idavidrein/gpqa
71
+ args:
72
+ num_few_shot: 0
73
+ metrics:
74
+ - type: acc_norm
75
+ value: 17.34
76
+ name: acc_norm
77
+ source:
78
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-Qwen-2.5-14B
79
+ name: Open LLM Leaderboard
80
+ - task:
81
+ type: text-generation
82
+ name: Text Generation
83
+ dataset:
84
+ name: MuSR (0-shot)
85
+ type: TAUR-Lab/MuSR
86
+ args:
87
+ num_few_shot: 0
88
+ metrics:
89
+ - type: acc_norm
90
+ value: 15.44
91
+ name: acc_norm
92
+ source:
93
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-Qwen-2.5-14B
94
+ name: Open LLM Leaderboard
95
+ - task:
96
+ type: text-generation
97
+ name: Text Generation
98
+ dataset:
99
+ name: MMLU-PRO (5-shot)
100
+ type: TIGER-Lab/MMLU-Pro
101
+ config: main
102
+ split: test
103
+ args:
104
+ num_few_shot: 5
105
+ metrics:
106
+ - type: acc
107
+ value: 45.47
108
+ name: accuracy
109
+ source:
110
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-Qwen-2.5-14B
111
+ name: Open LLM Leaderboard
112
  ---
113
 
114
  # Uploaded model
 
119
 
120
  This qwen2 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
121
 
122
+ [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
123
+ # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard)
124
+ Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_theprint__ReWiz-Qwen-2.5-14B)
125
+
126
+ | Metric |Value|
127
+ |-------------------|----:|
128
+ |Avg. |29.64|
129
+ |IFEval (0-Shot) |27.85|
130
+ |BBH (3-Shot) |44.86|
131
+ |MATH Lvl 5 (4-Shot)|26.89|
132
+ |GPQA (0-shot) |17.34|
133
+ |MuSR (0-shot) |15.44|
134
+ |MMLU-PRO (5-shot) |45.47|
135
+