usvsnsp commited on
Commit
adebe0a
1 Parent(s): 380366f

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +100 -0
README.md ADDED
@@ -0,0 +1,100 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ dataset_info:
3
+ features:
4
+ - name: sequence_id
5
+ dtype: int64
6
+ - name: text
7
+ dtype: string
8
+ - name: sequence_duplicates
9
+ dtype: int64
10
+ - name: max_frequency
11
+ dtype: int64
12
+ - name: avg_frequency
13
+ dtype: float64
14
+ - name: min_frequency
15
+ dtype: int64
16
+ - name: median_frequency
17
+ dtype: float64
18
+ - name: p25_frequency
19
+ dtype: int64
20
+ - name: p75_frequency
21
+ dtype: int64
22
+ - name: frequencies
23
+ sequence: int64
24
+ - name: is_incrementing
25
+ dtype: bool
26
+ - name: tokens
27
+ sequence: int64
28
+ - name: repeating_offset
29
+ dtype: int32
30
+ - name: num_repeating
31
+ dtype: int32
32
+ - name: smallest_repeating_chunk
33
+ sequence: int64
34
+ - name: memorization_score
35
+ dtype: float64
36
+ - name: templating_frequency_0.9
37
+ dtype: int64
38
+ - name: templating_frequency_0.8
39
+ dtype: int64
40
+ - name: prompt_perplexity
41
+ dtype: float32
42
+ - name: generation_perplexity
43
+ dtype: float32
44
+ - name: sequence_perplexity
45
+ dtype: float32
46
+ splits:
47
+ - name: pile.duped.70m
48
+ num_bytes: 7003348430
49
+ num_examples: 5000000
50
+ - name: pile.duped.160m
51
+ num_bytes: 7003348430
52
+ num_examples: 5000000
53
+ - name: pile.duped.410m
54
+ num_bytes: 7003348430
55
+ num_examples: 5000000
56
+ - name: pile.duped.1b
57
+ num_bytes: 7003348430
58
+ num_examples: 5000000
59
+ - name: pile.duped.1.4b
60
+ num_bytes: 7003348430
61
+ num_examples: 5000000
62
+ - name: pile.duped.2.8b
63
+ num_bytes: 7003348430
64
+ num_examples: 5000000
65
+ - name: pile.duped.6.9b
66
+ num_bytes: 7003348430
67
+ num_examples: 5000000
68
+ - name: pile.duped.12b
69
+ num_bytes: 7003348430
70
+ num_examples: 5000000
71
+ - name: pile.deduped.70m
72
+ num_bytes: 7013409756
73
+ num_examples: 5000000
74
+ - name: pile.deduped.160m
75
+ num_bytes: 7013409756
76
+ num_examples: 5000000
77
+ - name: pile.deduped.410m
78
+ num_bytes: 7013409756
79
+ num_examples: 5000000
80
+ - name: pile.deduped.1b
81
+ num_bytes: 7013409756
82
+ num_examples: 5000000
83
+ - name: pile.deduped.1.4b
84
+ num_bytes: 7013409756
85
+ num_examples: 5000000
86
+ - name: pile.deduped.2.8b
87
+ num_bytes: 7013409756
88
+ num_examples: 5000000
89
+ - name: pile.deduped.6.9b
90
+ num_bytes: 7013409756
91
+ num_examples: 5000000
92
+ - name: pile.deduped.12b
93
+ num_bytes: 7013409756
94
+ num_examples: 5000000
95
+ download_size: 48107269588
96
+ dataset_size: 112134065488
97
+ ---
98
+ # Dataset Card for "pile-semantic-memorization-filter-results"
99
+
100
+ [More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)