Datasets:

Modalities:
Text
Formats:
parquet
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
jlli commited on
Commit
425420e
1 Parent(s): 0840908

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -41,7 +41,7 @@ dataset_info:
41
  ### Dataset Summary
42
  xOA22 consists of 22 prompts originally shown in Appendix E, page 25 of the [OpenAssistant Conversations paper](https://arxiv.org/pdf/2304.07327.pdf). These 22 prompts were then manually translated by volunteers into 5 languages: Arabic, Simplified Chinese, French, Hindi and Spanish.
43
 
44
- These prompts were originally created for human evaluations of the multilingual abilities of BLOOMChat. Since not all prompts could be directly translatable due to cultural and linguistic differences, volunteers were encouraged to make appropriate substitutions and modifications that would maintain the intent of the original English prompt. As this was largely a collaborative, volunteer-led effort, this led to some discrepancies in the number of prompts per language. We make note of major departures from the original English prompt below.
45
  ### Languages
46
  - Arabic (ar)
47
  - English (en)
@@ -79,7 +79,7 @@ Below are notes from volunteer translators. Note that the Hindi split does not i
79
  - No translation notes
80
  - Hindi
81
  - Prompt 1: Replace "GLaDOS" with "Ravan", a famous antagonist from mythology
82
- - Prompt 4: This prompt was left untranslated, and so **is missing from the Hindi split**. Translator reasons are as follows: generaly people won't ask this in Hindi. Code writing community is generally English aware and they are most probably going to ask this question in English.
83
  - Prompt 9: Corresponds to English Prompt 10. Specific names changed to well known persons in Hindi speaking world.
84
  - Prompt 11: Corresponds to English Prompt 12. I removed "in depth crafting system", "directly or through emulation"
85
  - Prompt 21: Corresponds to English Prompt 22. I removed "social security numbers", "Google", and "Apple"
 
41
  ### Dataset Summary
42
  xOA22 consists of 22 prompts originally shown in Appendix E, page 25 of the [OpenAssistant Conversations paper](https://arxiv.org/pdf/2304.07327.pdf). These 22 prompts were then manually translated by volunteers into 5 languages: Arabic, Simplified Chinese, French, Hindi and Spanish.
43
 
44
+ These prompts were originally created for human evaluations of the multilingual abilities of [BLOOMChat](https://huggingface.co/sambanovasystems/BLOOMChat-176B-v1). Since not all prompts could be directly translatable due to cultural and linguistic differences, volunteers were encouraged to make appropriate substitutions and modifications that would maintain the intent of the original English prompt. As this was largely a collaborative, volunteer-led effort, this led to some discrepancies in the number of prompts per language. We make note of major departures from the original English prompt below.
45
  ### Languages
46
  - Arabic (ar)
47
  - English (en)
 
79
  - No translation notes
80
  - Hindi
81
  - Prompt 1: Replace "GLaDOS" with "Ravan", a famous antagonist from mythology
82
+ - Prompt 4: This prompt was left untranslated, and so **is missing from the Hindi split**. Translator reasons are as follows: generally people won't ask this in Hindi. Code writing community is generally English aware and they are most probably going to ask this question in English.
83
  - Prompt 9: Corresponds to English Prompt 10. Specific names changed to well known persons in Hindi speaking world.
84
  - Prompt 11: Corresponds to English Prompt 12. I removed "in depth crafting system", "directly or through emulation"
85
  - Prompt 21: Corresponds to English Prompt 22. I removed "social security numbers", "Google", and "Apple"