Datasets:
Tasks:
Text2Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10M - 100M
Tags:
flan
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -107,6 +107,13 @@ tags:
|
|
| 107 |
> [!IMPORTANT]
|
| 108 |
> each config has only been deduped **individually** thus far
|
| 109 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 110 |
|
| 111 |
deduped with:
|
| 112 |
|
|
|
|
| 107 |
> [!IMPORTANT]
|
| 108 |
> each config has only been deduped **individually** thus far
|
| 109 |
|
| 110 |
+
- all configs/subsets have columns `inputs` and `targets`
|
| 111 |
+
- deduped on `inputs`
|
| 112 |
+
- filtered for lang `en` if contents more than 5 chars.
|
| 113 |
+
- filter out any row with less than 1 char for either column
|
| 114 |
+
- `clean-text` applied to both columns
|
| 115 |
+
|
| 116 |
+
## dedup command
|
| 117 |
|
| 118 |
deduped with:
|
| 119 |
|