Hugging Face's logo

Datasets:
TabQueryBench2026
/
server

Dataset card Files
xet
 Community
1
TabQueryBench commited on
Commit
2db03b2
·
verified ·
1 Parent(s): e8da249

Add files using upload-large-folder tool

Browse files
Files changed (28) hide show
  1. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/_tabsyn_sample.py +43 -0
  2. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/_tabsyn_train.py +69 -0
  3. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_cat_test.npy +3 -0
  4. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_cat_train.npy +3 -0
  5. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_num_test.npy +3 -0
  6. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_num_train.npy +3 -0
  7. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/info.json +3 -0
  8. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/test.csv +3 -0
  9. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/train.csv +3 -0
  10. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/y_test.npy +3 -0
  11. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/y_train.npy +3 -0
  12. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/gen_20260504_051538.log +3 -0
  13. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/input_snapshot.json +3 -0
  14. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/public_gate/normalized_schema_snapshot.json +3 -0
  15. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/public_gate/public_gate_report.json +3 -0
  16. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/public_gate/staged_input_manifest.json +3 -0
  17. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/runtime_result.json +3 -0
  18. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/staged_features.json +3 -0
  19. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/test.csv +3 -0
  20. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/train.csv +3 -0
  21. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/val.csv +3 -0
  22. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/tabsyn/adapter_report.json +3 -0
  23. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/tabsyn/adapter_transforms_applied.json +3 -0
  24. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/tabsyn/model_input_manifest.json +3 -0
  25. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/synthetic/tabsyn_c14/real.csv +3 -0
  26. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/synthetic/tabsyn_c14/test.csv +3 -0
  27. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/tabsyn-c14-240000-20260504_051538.csv +3 -0
  28. syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/train_20260504_042709.log +3 -0
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/_tabsyn_sample.py ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os, sys, subprocess
2
+
3
+ work_dir = "/work/output-Benchmark-trainonly-v1/c14/tabsyn/tabsyn-c14-20260504_042655"
4
+ dataname = "tabsyn_c14"
5
+ output_csv = "/work/output-Benchmark-trainonly-v1/c14/tabsyn/tabsyn-c14-20260504_042655/tabsyn-c14-240000-20260504_051538.csv"
6
+ tabsyn_root = "/workspace/tabsyn"
7
+
8
+ assert os.path.exists(tabsyn_root), f"TabSyn source not mounted: {tabsyn_root}"
9
+
10
+ old = os.environ.get("PYTHONPATH", "")
11
+ os.environ["PYTHONPATH"] = tabsyn_root + (os.pathsep + old if old else "")
12
+ sys.path.insert(0, tabsyn_root)
13
+
14
+ os.chdir(tabsyn_root)
15
+
16
+ # Ensure data symlink exists
17
+ data_link = os.path.join(tabsyn_root, "data", dataname)
18
+ data_src = os.path.join(work_dir, "data", dataname)
19
+ os.makedirs(os.path.join(tabsyn_root, "data"), exist_ok=True)
20
+ if os.path.exists(data_link):
21
+ os.remove(data_link)
22
+ os.symlink(data_src, data_link)
23
+
24
+ print(f"[TabSyn] Sampling 240000 rows")
25
+ env = os.environ.copy()
26
+ env.setdefault("TABSYN_RESUME", "0")
27
+ env.setdefault("TABSYN_VAE_BATCH_SIZE", "32")
28
+ env.setdefault("TABSYN_VAE_EVAL_BATCH_SIZE", env["TABSYN_VAE_BATCH_SIZE"])
29
+ env.setdefault("TABSYN_VAE_INFER_BATCH_SIZE", env["TABSYN_VAE_BATCH_SIZE"])
30
+ env.setdefault("TABSYN_VAE_ENCODE_BATCH_SIZE", env["TABSYN_VAE_BATCH_SIZE"])
31
+ ret = subprocess.run(
32
+ [sys.executable, "main.py",
33
+ "--dataname", dataname,
34
+ "--mode", "sample",
35
+ "--method", "tabsyn",
36
+ "--gpu", "0",
37
+ "--save_path", output_csv],
38
+ cwd=tabsyn_root,
39
+ env=env
40
+ )
41
+ if ret.returncode != 0:
42
+ sys.exit(ret.returncode)
43
+ print(f"[TabSyn] Saved -> {output_csv}")
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/_tabsyn_train.py ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os, sys, subprocess
2
+
3
+ work_dir = "/work/output-Benchmark-trainonly-v1/c14/tabsyn/tabsyn-c14-20260504_042655"
4
+ dataname = "tabsyn_c14"
5
+ tabsyn_root = "/workspace/tabsyn"
6
+
7
+ assert os.path.exists(tabsyn_root), f"TabSyn source not mounted: {tabsyn_root}"
8
+
9
+ old = os.environ.get("PYTHONPATH", "")
10
+ os.environ["PYTHONPATH"] = tabsyn_root + (os.pathsep + old if old else "")
11
+ sys.path.insert(0, tabsyn_root)
12
+
13
+ os.chdir(tabsyn_root)
14
+
15
+ # Symlink data dir into TabSyn data/
16
+ data_link = os.path.join(tabsyn_root, "data", dataname)
17
+ data_src = os.path.join(work_dir, "data", dataname)
18
+ os.makedirs(os.path.join(tabsyn_root, "data"), exist_ok=True)
19
+ if os.path.exists(data_link):
20
+ os.remove(data_link)
21
+ os.symlink(data_src, data_link)
22
+
23
+ env = os.environ.copy()
24
+ env.setdefault("TABSYN_RESUME", "0")
25
+ env.setdefault("TABSYN_VAE_BATCH_SIZE", "32")
26
+ env.setdefault("TABSYN_VAE_NUM_WORKERS", "0")
27
+ env.setdefault("TABSYN_VAE_EVAL_BATCH_SIZE", env["TABSYN_VAE_BATCH_SIZE"])
28
+ env.setdefault("TABSYN_VAE_INFER_BATCH_SIZE", env["TABSYN_VAE_BATCH_SIZE"])
29
+ env.setdefault("TABSYN_VAE_ENCODE_BATCH_SIZE", env["TABSYN_VAE_BATCH_SIZE"])
30
+ # Safer defaults for wide tables on Docker: reduce shared-memory pressure in diffusion DataLoader.
31
+ env.setdefault("TABSYN_DIFFUSION_NUM_WORKERS", "0")
32
+ _te = 10
33
+ if _te is not None:
34
+ env["TABSYN_VAE_EPOCHS"] = str(_te)
35
+ env["TABSYN_DIFFUSION_MAX_EPOCHS"] = str(max(_te + 1, 2))
36
+
37
+ # Data preprocessing is done on the host side (_prepare_data_dir)
38
+ # which creates .npy files, train/test CSVs, and info.json
39
+
40
+ # Step 1: Train VAE (produces latent embeddings)
41
+ print(f"[TabSyn] Step 1/2: Training VAE in {tabsyn_root}, dataname={dataname}")
42
+ ret = subprocess.run(
43
+ [sys.executable, "main.py",
44
+ "--dataname", dataname,
45
+ "--mode", "train",
46
+ "--method", "vae",
47
+ "--gpu", "0"],
48
+ cwd=tabsyn_root,
49
+ env=env
50
+ )
51
+ if ret.returncode != 0:
52
+ print("[TabSyn] VAE training failed")
53
+ sys.exit(ret.returncode)
54
+
55
+ # Step 2: Train diffusion model on latent space
56
+ print(f"[TabSyn] Step 2/2: Training diffusion model")
57
+ ret = subprocess.run(
58
+ [sys.executable, "main.py",
59
+ "--dataname", dataname,
60
+ "--mode", "train",
61
+ "--method", "tabsyn",
62
+ "--gpu", "0"],
63
+ cwd=tabsyn_root,
64
+ env=env
65
+ )
66
+ if ret.returncode != 0:
67
+ print("[TabSyn] Diffusion training failed")
68
+ sys.exit(ret.returncode)
69
+ print("[TabSyn] Training complete (VAE + Diffusion)")
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_cat_test.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96fb5a86a8b5c99f151626dee541d9024bac427c393b9ed577c13dbcc6b6832d
3
+ size 38400128
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_cat_train.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96fb5a86a8b5c99f151626dee541d9024bac427c393b9ed577c13dbcc6b6832d
3
+ size 38400128
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_num_test.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3026532dc1e8f61b12f8f829ec618733e69f9da70d40e8234069ff10f3381878
3
+ size 3840128
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/X_num_train.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3026532dc1e8f61b12f8f829ec618733e69f9da70d40e8234069ff10f3381878
3
+ size 3840128
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/info.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1abb4aa0eddc6c8e155d50346f6445109d3a84210396de9ae8386f98db8dd709
3
+ size 3677
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/test.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5921345765f8af3133fc50bf596097efd9e2f166a6a923b4951b080daf112c4
3
+ size 16124837
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/train.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5921345765f8af3133fc50bf596097efd9e2f166a6a923b4951b080daf112c4
3
+ size 16124837
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/y_test.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9feaa3857a64a11ee3e12f755d82c94f0e23fcdd5f916c943a48340cb510db33
3
+ size 1920128
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/data/tabsyn_c14/y_train.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9feaa3857a64a11ee3e12f755d82c94f0e23fcdd5f916c943a48340cb510db33
3
+ size 1920128
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/gen_20260504_051538.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e9011555b272d0fda9bae19a1c98bcfb693ea36c19693c9ff0a40a54b1fdfb6c
3
+ size 955
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/input_snapshot.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1af6bd18f0bb7fb5827d89a2bb194fff828a10a6c80ba9b28b2a212382187ea6
3
+ size 1363
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/public_gate/normalized_schema_snapshot.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e559ea7c300bee8dca838c3419b8517cfe17c153470e56a6f1c1fab72b7a10b
3
+ size 11306
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/public_gate/public_gate_report.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:465dc423a1ca46fea5f39f1b0c83b6cce2c16cdb3be3f582e8ce20fbce674981
3
+ size 920
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/public_gate/staged_input_manifest.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a7edcd7f67fdf5ac4a692d212ac9e0fbb24a3725e5389b8955e795d615bafb9
3
+ size 12112
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/runtime_result.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97290acc41ac03c03925f9706b22ba56e0ac9ba8c1783a3abb12808ab02ffae9
3
+ size 885
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/staged_features.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c92cf6530ea76cef171daa9d723513c36aafe5dc5d8f5dfc27778090e2b0c03
3
+ size 2288
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/test.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:617ecb620416375c67546e0fa5a8b9a3923d689bda904dc23824ea175d1f8597
3
+ size 3965839
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/train.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12398ccb68a2499b9c126f61b25f03c36a65f026d755e1d5ff5653e614676167
3
+ size 31717079
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/public/val.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f40dbd7bbd9e2783c24d0d12ec35a85581835d45238774eb0676de44a734feda
3
+ size 3965919
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/tabsyn/adapter_report.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c76ab6c0adf650e314be8cdcf782cce333a846d6e9cddd649776ee3ffb261ce2
3
+ size 320
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/tabsyn/adapter_transforms_applied.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f53cda18c2baa0c0354bb5f9a3ecbe5ed12ab4d8e11ba873c2f11161202b945
3
+ size 2
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/staged/tabsyn/model_input_manifest.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fdd39a9aa1adaaf282d9917a5c0cbf3256863967fd5439be08995cc69d733b1c
3
+ size 12308
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/synthetic/tabsyn_c14/real.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5921345765f8af3133fc50bf596097efd9e2f166a6a923b4951b080daf112c4
3
+ size 16124837
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/synthetic/tabsyn_c14/test.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5921345765f8af3133fc50bf596097efd9e2f166a6a923b4951b080daf112c4
3
+ size 16124837
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/tabsyn-c14-240000-20260504_051538.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24f2d9ff4ae9d67bb7f82aadfc89c85655dc78debc270f553187c061c92e97ad
3
+ size 18392001
syntheticSuccess/c14/tabsyn/tabsyn-c14-20260504_042655/train_20260504_042709.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00c438a488c258b5b796279d36fcf3b827ce378fc49b7ccdfa1b0d05f4e3d5dc
3
+ size 1335790