Datasets:

Languages:
English
ArXiv:
Tags:
math
hails commited on
Commit
d94c6a0
1 Parent(s): 89c96f9

Add explicit filenames as workaround

Browse files
Files changed (1) hide show
  1. proof-pile-2.py +79 -3
proof-pile-2.py CHANGED
@@ -19,6 +19,82 @@ A dataset of high quality mathematical text. """
19
  _HOMEPAGE = "https://github.com/EleutherAI/math-lm"
20
 
21
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  class ProofPile2Config(datasets.BuilderConfig):
23
  """BuilderConfig for RedPajama sample."""
24
 
@@ -97,9 +173,9 @@ class ProofPile2(datasets.GeneratorBasedBuilder):
97
  "data_files": list(map(
98
  dl_manager.download,
99
  [
100
- x
101
- for subset in self.config.subsets
102
- for x in dl_manager.iter_files("https://huggingface.co/datasets/EleutherAI/proof-pile-2/tree/main/{subset}/{split}")
103
  ]
104
  ))
105
  },
 
19
  _HOMEPAGE = "https://github.com/EleutherAI/math-lm"
20
 
21
 
22
+ # hacky workaround: listing out files here for download, because removing dataloader script entirely introduced another bug
23
+ _ARXIV_FILES = {
24
+ "train": [f"arxiv_{i:03}.jsonl.zst" for i in range(100)],
25
+ "validation": [f"arxiv_{i:03}.jsonl.zst" for i in range(100)],
26
+ "test": [f"arxiv_{i:03}.jsonl.zst" for i in range(100)],
27
+ }
28
+ _OWM_FILES = {
29
+ "train": [f"shard-{i:04}.jsonl.zst" for i in range(63)],
30
+ "validation": ["val.jsonl.zst"],
31
+ "test": ["test.jsonl.zst"],
32
+ }
33
+ _ALGSTACK_FILES = {
34
+ "train": ["agda0000.jsonl.zst", "c000.jsonl.zst"]
35
+ + [f"cpp{i:04}.jsonl.zst" for i in range(5)]
36
+ + [f"fortran{i:04}.jsonl.zst" for i in range(4)]
37
+ + ["gap0000.jsonl.zst"]
38
+ + [f"github-MATLAB-train-{i:04}.jsonl.zst" for i in range(4)]
39
+ + [f"github-coq-train-{i:04}.jsonl.zst" for i in range(3)]
40
+ + ["github-isabelle-train-0000.jsonl.zst", "github-lean-train-0000.jsonl.zst"]
41
+ + ["haskell0000.jsonl.zst", "idris0000.jsonl.zst", "isa_proofsteps.jsonl.zst"]
42
+ + [f"julia{i:04}.jsonl.zst" for i in range(6)]
43
+ + ["jupyter-notebook0000.jsonl.zst", "lean_proofsteps.jsonl.zst", "maple0000.jsonl.zst"]
44
+ + [f"python{i:04}.jsonl.zst" for i in range(42)]
45
+ + ["r0000.jsonl.szt"]
46
+ + [f"tex{i:04}.jsonl.zst" for i in range(3)],
47
+ "validation": [
48
+ "agda-validation.jsonl.zst",
49
+ "c-validation.jsonl.zst",
50
+ "cpp-validation.jsonl.zst",
51
+ "fortran-validation.jsonl.zst",
52
+ "gap-validation.jsonl.zst",
53
+ "github-MATLAB-validation-0000.jsonl.zst",
54
+ "github-coq-validation-0000.jsonl.zst",
55
+ "github-isabelle-validation-0000.jsonl.zst",
56
+ "github-lean-validation-0000.jsonl.zst",
57
+ "haskell-validation.jsonl.zst",
58
+ "idris-validation.jsonl.zst",
59
+ "isa_proofsteps.jsonl.zst",
60
+ "julia-validation.jsonl.zst",
61
+ "jupyter-notebook-validation.jsonl.zst",
62
+ "lean_proofsteps.jsonl.zst",
63
+ "maple-validation.jsonl.zst",
64
+ "python-validation.jsonl.zst",
65
+ "r-validation.jsonl.zst",
66
+ "tex-validation.jsonl.zst",
67
+ ],
68
+ "test": [
69
+ "agda-test.jsonl.zst",
70
+ "c-test.jsonl.zst",
71
+ "cpp-test.jsonl.zst",
72
+ "fortran-test.jsonl.zst",
73
+ "gap-test.jsonl.zst",
74
+ "github-MATLAB-test-0000.jsonl.zst",
75
+ "github-coq-test-0000.jsonl.zst",
76
+ "github-isabelle-test-0000.jsonl.zst",
77
+ "github-lean-test-0000.jsonl.zst",
78
+ "haskell-test.jsonl.zst",
79
+ "idris-test.jsonl.zst",
80
+ "isa_proofsteps.jsonl.zst",
81
+ "julia-test.jsonl.zst",
82
+ "jupyter-notebook-test.jsonl.zst",
83
+ "lean_proofsteps.jsonl.zst",
84
+ "maple-test.jsonl.zst",
85
+ "python-test.jsonl.zst",
86
+ "r-test.jsonl.zst",
87
+ "tex-test.jsonl.zst",
88
+ ]
89
+ }
90
+
91
+ _FILES_MAPPING = {
92
+ "arxiv": _ARXIV_FILES,
93
+ "open-web-math": _OWM_FILES,
94
+ "algebraic-stack": _ALGSTACK_FILES,
95
+ }
96
+
97
+
98
  class ProofPile2Config(datasets.BuilderConfig):
99
  """BuilderConfig for RedPajama sample."""
100
 
 
173
  "data_files": list(map(
174
  dl_manager.download,
175
  [
176
+ f"https://huggingface.co/datasets/EleutherAI/proof-pile-2/tree/main/{subset}/{split}/{x}"
177
+ for subset in self.config.subsets
178
+ for x in _FILES_MAPPING[subset][split]
179
  ]
180
  ))
181
  },