artms007 commited on
Commit
8c806c7
1 Parent(s): 31c3cd7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -3
README.md CHANGED
@@ -37,9 +37,9 @@ size_categories:
37
  ## Dataset Summary
38
  This is a Japanese translated version of HumanEval, an evaluation harness for the HumanEval problem solving dataset described in the paper "Evaluating Large Language Models Trained on Code".
39
 
40
- LLMのコード生成能力の標準ベンチマーク HumanEval の日本語翻訳版です。
41
- 機械翻訳(DeepL, GPT-4)の翻訳結果を全て人手によって再修正し、 訳文を日本人のプログラマが読んで理解し、コードが書ける内容かチェックしました。
42
- ただし、英語版 HumanEval の間違いは、修正せずに残して、 HumanEval 同様に不完全なドキュメントからの生成能力を見るようになっています。
43
  日本語LLMのベンチマークとしてお使いください。
44
 
45
  ## Languages
@@ -62,6 +62,7 @@ DatasetDict({
62
  ### Data Instances
63
  An example of a dataset instance:
64
 
 
65
  {
66
  "task_id": "test/0",
67
  "prompt_en": "def return1():\n \"\"\"\n A simple function that returns the integer 1.\n \"\"\"\n",
@@ -70,6 +71,7 @@ An example of a dataset instance:
70
  "test": "def check(candidate):\n assert candidate() == 1",
71
  "entry_point": "return1"
72
  }
 
73
 
74
  ### Data Fields
75
  - task_id: Unique identifier for a task.
 
37
  ## Dataset Summary
38
  This is a Japanese translated version of HumanEval, an evaluation harness for the HumanEval problem solving dataset described in the paper "Evaluating Large Language Models Trained on Code".
39
 
40
+ LLMのコード生成能力の標準ベンチマーク HumanEval の日本語翻訳版です。
41
+ 機械翻訳(DeepL, GPT-4)の翻訳結果を全て人手によって再修正し、 訳文を日本人のプログラマが読んで理解し、コードが書ける内容かチェックしました。
42
+ ただし、英語版 HumanEval の間違いは、修正せずに残して、 HumanEval 同様に不完全なドキュメントからの生成能力を見るようになっています。
43
  日本語LLMのベンチマークとしてお使いください。
44
 
45
  ## Languages
 
62
  ### Data Instances
63
  An example of a dataset instance:
64
 
65
+ ```
66
  {
67
  "task_id": "test/0",
68
  "prompt_en": "def return1():\n \"\"\"\n A simple function that returns the integer 1.\n \"\"\"\n",
 
71
  "test": "def check(candidate):\n assert candidate() == 1",
72
  "entry_point": "return1"
73
  }
74
+ ```
75
 
76
  ### Data Fields
77
  - task_id: Unique identifier for a task.