Adamiros commited on
Commit
ed3e7c6
·
verified ·
1 Parent(s): 42ff59f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -46
README.md CHANGED
@@ -7,49 +7,4 @@ sdk: static
7
  pinned: false
8
  ---
9
 
10
- Outer pipes Cell padding
11
- No sorting
12
- | Benchmark | Metric [%] | Gemma2-2b-it | QWEN-2-1.5b | GPT-3.5 Turbo | Claude3 Haiku | Llama3 8B | Gemma2 9B It | Gemma2 27B It |
13
- | --------------------------- | ---------- | ------------:| -----------:| -------------:| -------------:| ---------:| ------------:| -------------:|
14
- | AGREE | ACC | 43.86 | 37.16 | 46.7 | 65.7 | 32.5 | 55.02 | 59.58 |
15
- | ANLI | ACC | 39.58 | 36.67 | 44.7 | 51.5 | 43 | 57.17 | 61.33 |
16
- | | F1 | 25.39 | 34.53 | 41.9 | 50.8 | 39.2 | 57.32 | 61.32 |
17
- | ANLI EN | ACC | 41.33 | 39.92 | 44.25 | 55.34 | 46.8 | 55 | |
18
- | | F1 | 38.26 | 33.19 | 40.58 | 54.16 | 38.8 | 52.97 | |
19
- | ARC Challenge | ACC | 52.79 | 44.66 | 73.1 | 76.8 | 64.3 | 84.56 | 88.14 |
20
- | ARC Challenge EN | ACC | 72.37 | 69.75 | 82.92 | 77.95 | 78.5 | 90.61 | |
21
- | ARC Easy | ACC | 69.94 | | 85.8 | 85.3 | 79.4 | 92.28 | 95.01 |
22
- | ARC Easy EN | ACC | 87.15 | 85.59 | 93.1 | 89.1 | 91.4 | 96.63 | |
23
- | Belebele | ACC | 69.02 | 55.42 | 80.3 | 88.2 | 76.5 | 90.38 | 89.16 |
24
- | Belebele EN | ACC | 79.33 | 70.84 | 87.04 | 90.95 | 83.4 | 94.06 | |
25
- | CTKFacts | ACC | 56.63 | 55.56 | 61.8 | 69.6 | 61.7 | 64.34 | 70.79 |
26
- | | F1 | 45.9 | 48.98 | 47.7 | 62 | 51.6 | 59.05 | 68.44 |
27
- | CTKFacts EN | ACC | 46.95 | 50 | 67.56 | 68.06 | 69 | 70.97 | |
28
- | | F1 | 44.56 | 39.94 | 63.23 | 62.22 | 65 | 69.32 | |
29
- | Czech News | ACC | 62.9 | 17.8 | 78.9 | 81.3 | 71.6 | 79.5 | 82.55 |
30
- | | F1 | 60.5 | 14.83 | 78.5 | 81.3 | 70.7 | 78.9 | 81.64 |
31
- | Facebook Comments | ACC | 70.7 | 52.8 | 71.5 | 75.8 | 66.8 | 75.2 | 75.24\* |
32
- | | F1 | 69.73 | 46.79 | 69 | 74.1 | 64.1 | 74.51 | 73.78\* |
33
- | GSM8K | ACC | 28.84 | 18.29 | 64.2 | 78.6 | 67.07\* | \- | 75.71\* |
34
- | GSM8K EN | ACC | 66.89 | 54.59 | 83.14 | 88.98 | \- | 87.96 | |
35
- | Klokánek | ACC | 23.51 | 21.29 | 29.3 | 24.5 | 21.8 | 29.6 | 28.93 |
36
- | Mall Reviews | ACC | 62.1 | 48.87 | 59.8 | 57.7 | 59.5 | 60.93 | 61.03 |
37
- | | F1 | 60.13 | 43.89 | 55.4 | 55.2 | 57.3 | 59.88 | 60.38 |
38
- | MMLU | ACC | 43.62 | 39.86 | 58 | 67.3 | 46.8 | 54.49 | 68.07 |
39
- | MMLU EN | ACC | 56.82 | 53.04 | 64.89 | 73.01 | 53.5 | 49.84 | |
40
- | SNLI | ACC | 57.44 | 59.06 | 61.8 | 71.7 | 60.9 | 77.56 | |
41
- | | F1 | 57.48 | 58.41 | 51.5 | 70.5 | 58.7 | 77.76 | |
42
- | SNLI EN | ACC | 61.61 | 48.74 | 60.57 | 72.74 | 65.5 | 78.29 | |
43
- | | F1 | 46.59 | 30.85 | 43.32 | 53.78 | 47.4 | 59.09 | |
44
- | SQAD | EM Acc | 67.85 | 58.84 | 66.2 | 59.8 | 67.6 | 78.29 | |
45
- | | BoW F1 | 79.18 | 68.83 | 83.5 | 76.3 | 82.5 | 88.77 | |
46
- | SQuAD (Generation) | EM Acc | 34.05 | 16.9 | 37.3 | 36.3 | 36.6 | 52.52 | |
47
- | | BoW F1 | 39.29 | 22.04 | 43 | 44.7 | 44.5 | 57.87 | |
48
- | SQuAD (No-Answer Detection) | EM Acc | 56.2 | 45.75 | 52.4 | 60.3 | 56.8 | 64.88 | |
49
- | | BoW F1 | 50.76 | 31.67 | 44.2 | 56.4 | 51.3 | 62.47 | |
50
- | Subjectivity | ACC | 68.9 | 50.95 | 80.2 | 81.5 | 78.3 | 85.9 | |
51
- | | F1 | 66.79 | 36.62 | 80.2 | 81.2 | 77.6 | 85.68 | |
52
- | Subjectivity EN | ACC | 77.6 | 53.1 | 86.8 | 86.6 | 87.1 | 86 | |
53
- | | F1 | 76.7 | 53.05 | 86.79 | 86.59 | 87.1 | 85.85 | |
54
- | TruthfulQA | ACC | 45.81 | 31.65 | 53.5 | 65.8 | 41 | 62.62 | |
55
- | TruthfulQA EN | ACC | 49.75 | 32.02 | 58.5 | 70.81 | 41.6 | 68.06 |
 
7
  pinned: false
8
  ---
9
 
10
+ Placeholder README