rombodawg commited on
Commit
fc5807b
1 Parent(s): 9604a24

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +127 -3
README.md CHANGED
@@ -1,3 +1,127 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ library_name: transformers
4
+ base_model:
5
+ - Qwen/Qwen2.5-14B-Instruct
6
+ model-index:
7
+ - name: Replete-LLM-V2.5-Qwen-14b
8
+ results:
9
+ - task:
10
+ type: text-generation
11
+ name: Text Generation
12
+ dataset:
13
+ name: IFEval (0-Shot)
14
+ type: HuggingFaceH4/ifeval
15
+ args:
16
+ num_few_shot: 0
17
+ metrics:
18
+ - type: inst_level_strict_acc and prompt_level_strict_acc
19
+ value: 58.4
20
+ name: strict accuracy
21
+ source:
22
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Replete-AI/Replete-LLM-V2.5-Qwen-14b
23
+ name: Open LLM Leaderboard
24
+ - task:
25
+ type: text-generation
26
+ name: Text Generation
27
+ dataset:
28
+ name: BBH (3-Shot)
29
+ type: BBH
30
+ args:
31
+ num_few_shot: 3
32
+ metrics:
33
+ - type: acc_norm
34
+ value: 49.39
35
+ name: normalized accuracy
36
+ source:
37
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Replete-AI/Replete-LLM-V2.5-Qwen-14b
38
+ name: Open LLM Leaderboard
39
+ - task:
40
+ type: text-generation
41
+ name: Text Generation
42
+ dataset:
43
+ name: MATH Lvl 5 (4-Shot)
44
+ type: hendrycks/competition_math
45
+ args:
46
+ num_few_shot: 4
47
+ metrics:
48
+ - type: exact_match
49
+ value: 15.63
50
+ name: exact match
51
+ source:
52
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Replete-AI/Replete-LLM-V2.5-Qwen-14b
53
+ name: Open LLM Leaderboard
54
+ - task:
55
+ type: text-generation
56
+ name: Text Generation
57
+ dataset:
58
+ name: GPQA (0-shot)
59
+ type: Idavidrein/gpqa
60
+ args:
61
+ num_few_shot: 0
62
+ metrics:
63
+ - type: acc_norm
64
+ value: 16.22
65
+ name: acc_norm
66
+ source:
67
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Replete-AI/Replete-LLM-V2.5-Qwen-14b
68
+ name: Open LLM Leaderboard
69
+ - task:
70
+ type: text-generation
71
+ name: Text Generation
72
+ dataset:
73
+ name: MuSR (0-shot)
74
+ type: TAUR-Lab/MuSR
75
+ args:
76
+ num_few_shot: 0
77
+ metrics:
78
+ - type: acc_norm
79
+ value: 18.83
80
+ name: acc_norm
81
+ source:
82
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Replete-AI/Replete-LLM-V2.5-Qwen-14b
83
+ name: Open LLM Leaderboard
84
+ - task:
85
+ type: text-generation
86
+ name: Text Generation
87
+ dataset:
88
+ name: MMLU-PRO (5-shot)
89
+ type: TIGER-Lab/MMLU-Pro
90
+ config: main
91
+ split: test
92
+ args:
93
+ num_few_shot: 5
94
+ metrics:
95
+ - type: acc
96
+ value: 48.62
97
+ name: accuracy
98
+ source:
99
+ url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Replete-AI/Replete-LLM-V2.5-Qwen-14b
100
+ name: Open LLM Leaderboard
101
+ ---
102
+ # Rombos-LLM-V2.5-Qwen-14b
103
+
104
+ ![image/png](https://cdn-uploads.huggingface.co/production/uploads/642cc1c253e76b4c2286c58e/ihnWXDEgV-ZKN_B036U1J.png)
105
+
106
+ Rombos-LLM-V2.5-Qwen-14b is a continues finetuned version of Qwen2.5-14B. I noticed recently that the Qwen team did not learn from my methods of continuous finetuning, the great benefits, and no downsides of it. So I took it upon myself to merge the instruct model with the base model myself using the *Ties* merge method
107
+
108
+ This version of the model shows higher performance than the original instruct and base models.
109
+
110
+ Quants:
111
+
112
+ GGUF: https://huggingface.co/bartowski/Replete-LLM-V2.5-Qwen-14b-GGUF
113
+
114
+ Benchmarks:
115
+ # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard)
116
+ Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_Replete-AI__Replete-LLM-V2.5-Qwen-14b)
117
+
118
+ | Metric |Value|
119
+ |-------------------|----:|
120
+ |Avg. |34.52|
121
+ |IFEval (0-Shot) |58.40|
122
+ |BBH (3-Shot) |49.39|
123
+ |MATH Lvl 5 (4-Shot)|15.63|
124
+ |GPQA (0-shot) |16.22|
125
+ |MuSR (0-shot) |18.83|
126
+ |MMLU-PRO (5-shot) |48.62|
127
+