TabQueryBench commited on
Commit
6a4f7db
·
verified ·
1 Parent(s): 5e10fd4

Create SynthData0523 metadata and provenance index

Browse files
SynthData0523/INDEX.csv ADDED
The diff for this file is too large to render. See raw diff
 
SynthData0523/INDEX.md ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # SynthData0523 Index Summary
2
+
3
+ ## Main
4
+
5
+ - Unique files: `21474`
6
+ - Unique runs: `656`
7
+ - Datasets: `49`
8
+ - Models: `12`
9
+
10
+ ### Duplicate-source run groups in `main/`
11
+
12
+ - `5090-Success`: `5` runs
13
+ - `5090-Success|syntheticSuccess`: `3` runs
14
+ - `syntheticSuccess`: `490` runs
15
+ - `syntheticSuccess|timecost`: `100` runs
16
+ - `timecost`: `58` runs
17
+
18
+ ## Hyper Parameter Tuning
19
+
20
+ - Files: `3220`
21
+ - Runs under `runs/`: `96`
22
+ - Datasets: `1`
23
+ - Models: `4`
24
+
25
+ ## Dataset Coverage
26
+
27
+ ### main/
28
+ - `c10`: `442` files
29
+ - `c11`: `266` files
30
+ - `c12`: `186` files
31
+ - `c13`: `551` files
32
+ - `c14`: `535` files
33
+ - `c15`: `507` files
34
+ - `c16`: `401` files
35
+ - `c17`: `418` files
36
+ - `c18`: `276` files
37
+ - `c19`: `416` files
38
+ - `c2`: `1347` files
39
+ - `c20`: `288` files
40
+ - `c3`: `265` files
41
+ - `c4`: `272` files
42
+ - `c5`: `265` files
43
+ - `c6`: `980` files
44
+ - `c7`: `466` files
45
+ - `c8`: `264` files
46
+ - `c9`: `244` files
47
+ - `m1`: `262` files
48
+ - `m10`: `262` files
49
+ - `m11`: `398` files
50
+ - `m12`: `286` files
51
+ - `m2`: `302` files
52
+ - `m4`: `1178` files
53
+ - `m5`: `373` files
54
+ - `m6`: `426` files
55
+ - `m7`: `315` files
56
+ - `m8`: `707` files
57
+ - `m9`: `394` files
58
+ - `n1`: `449` files
59
+ - `n10`: `296` files
60
+ - `n11`: `730` files
61
+ - `n12`: `361` files
62
+ - `n14`: `380` files
63
+ - `n15`: `487` files
64
+ - `n16`: `482` files
65
+ - `n17`: `366` files
66
+ - `n18`: `362` files
67
+ - `n19`: `488` files
68
+ - `n2`: `295` files
69
+ - `n20`: `219` files
70
+ - `n3`: `1179` files
71
+ - `n4`: `300` files
72
+ - `n5`: `297` files
73
+ - `n6`: `484` files
74
+ - `n7`: `364` files
75
+ - `n8`: `347` files
76
+ - `n9`: `296` files
77
+
78
+ ### hyper_parameter_tuning/
79
+ - `n11`: `3220` files
SynthData0523/README.md ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # SynthData0523
2
+
3
+ `SynthData0523` is a copied, reorganized view of valid synthetic generation assets as of 2026-05-23.
4
+
5
+ ## Structure
6
+
7
+ ```text
8
+ SynthData0523/
9
+ README.md
10
+ INDEX.csv
11
+ INDEX.md
12
+ main/
13
+ <dataset_id>/
14
+ <model_id>/
15
+ <run_id>/
16
+ hyper_parameter_tuning/
17
+ <dataset_id>/
18
+ <model_id>/
19
+ _rf_tuning/
20
+ runs/
21
+ <run_id>/
22
+ ```
23
+
24
+ ## Included Sources
25
+
26
+ - `main/` is built from `syntheticSuccess/`, `timecost/`, and `5090-Success/`.
27
+ - `hyper_parameter_tuning/` is copied from `Hyper ParameterTuning/` while preserving both `runs/` and `_rf_tuning/`.
28
+ - Original source directories are intentionally preserved and are not deleted.
29
+
30
+ ## Provenance Rules
31
+
32
+ - When the same run already exists in multiple source roots, `SynthData0523/main/` keeps one copied target path and records all source roots in `INDEX.csv`.
33
+ - Preferred copy source precedence for duplicate main assets is:
34
+ 1. `syntheticSuccess`
35
+ 2. `timecost`
36
+ 3. `5090-Success`
37
+ - `INDEX.csv` records file-level provenance for every copied file.
38
+
39
+ ## Current Snapshot
40
+
41
+ - `main/` unique files: `21474`
42
+ - `main/` unique runs: `656`
43
+ - `main/` datasets: `49`
44
+ - `main/` models: `12`
45
+ - `hyper_parameter_tuning/` files: `3220`
46
+ - `hyper_parameter_tuning/` runs: `96`
47
+
48
+ ## Notes
49
+
50
+ - `INDEX.md` provides a compact summary.
51
+ - `INDEX.csv` is the canonical machine-readable provenance index.