Datasets:

Modalities:
Text
Formats:
parquet
Languages:
English
Libraries:
Datasets
pandas
License:
arjunguha commited on
Commit
fa40fdc
1 Parent(s): 6d6272d

More README

Browse files
Files changed (1) hide show
  1. README.md +24 -1
README.md CHANGED
@@ -39,6 +39,29 @@ generation that supports 18 programming languages. It takes the OpenAI
39
  translate them to other languages. It is easy to add support for new languages
40
  and benchmarks.
41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
42
  ## Example
43
 
44
  The following script uses the Salesforce/codegen model to generate Lua
@@ -72,7 +95,7 @@ for problem in problems["test"]:
72
  return_tensors="pt",
73
  ).input_ids.cuda()
74
  generated_ids = model.generate(
75
- input_ids, max_length=256, pad_token_id=tokenizer.eos_token_id + 2
76
  )
77
  truncated_string = stop_at_stop_token(tokenizer.decode(generated_ids[0]), problem)
78
  filename = problem["name"] + "." + LANG
 
39
  translate them to other languages. It is easy to add support for new languages
40
  and benchmarks.
41
 
42
+ ## Subsets
43
+
44
+ For most purposes, you should use the variations called *SRCDATA-LANG*, where
45
+ *SRCDATA* is either "humaneval" or "mbpp" and *LANG* is one of the supported
46
+ languages. We use the canonical file extension for each language to identify
47
+ the language, e.g., "py" for Python, "cpp" for C++, "lua" for Lua, and so on.
48
+
49
+ We also provide a few other variations:
50
+
51
+ - *SRCDATA-LANG-keep* is the same as *SRCDATA-LANG*, but the text of the prompt
52
+ is totally unchanged. If the original prompt had Python doctests, they remain
53
+ as Python instead of being translated to *LANG*. If the original prompt had
54
+ Python-specific terminology, e.g., "list", it remains "list", instead of
55
+ being translated, e.g., to "vector" for C++.
56
+
57
+ - *SRCDATA-LANG-transform* transforms the doctests to *LANG* but leaves
58
+ the natural language text of the prompt unchanged.
59
+
60
+ - *SRCDATA-LANG-removed* removes the doctests from the prompt.
61
+
62
+ Note that MBPP does not have any doctests, so the "removed" and "transform"
63
+ variations are not available for MBPP.
64
+
65
  ## Example
66
 
67
  The following script uses the Salesforce/codegen model to generate Lua
 
95
  return_tensors="pt",
96
  ).input_ids.cuda()
97
  generated_ids = model.generate(
98
+ input_ids, max_length=512, pad_token_id=tokenizer.eos_token_id + 2
99
  )
100
  truncated_string = stop_at_stop_token(tokenizer.decode(generated_ids[0]), problem)
101
  filename = problem["name"] + "." + LANG