Update README.md
Browse files
README.md
CHANGED
|
@@ -43,4 +43,41 @@ tags:
|
|
| 43 |
- sentiment-classification
|
| 44 |
size_categories:
|
| 45 |
- 10K<n<100K
|
| 46 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
- sentiment-classification
|
| 44 |
size_categories:
|
| 45 |
- 10K<n<100K
|
| 46 |
+
---
|
| 47 |
+
|
| 48 |
+
# IndicVarna for Callchimp.ai (a Dynopii product)
|
| 49 |
+
|
| 50 |
+
We introduce IndiVarna which was prepared by using Google Translate the [dair-ai/emotion](https://huggingface.co/datasets/dair-ai/emotion) dataset to top 10 most commonly used Indian languages.
|
| 51 |
+
|
| 52 |
+
This dataset contains `10000` samples of each of the 10 languages supported.
|
| 53 |
+
|
| 54 |
+
The dataset further translated the labels in the dataset to 3 label sentiments - 0: Negative, 1: Neutral and 2: Positive. Each language has `3334` samples of each category of sentiment.
|
| 55 |
+
|
| 56 |
+
The dataset has three columns - `uuid`, `text`, `label`. `uuid` here is the concatenation of `uuid` with `language code` separated by `-`.
|
| 57 |
+
|
| 58 |
+
This enabled us to start using the dataset out of the box with Hugging Face based text classification pipelines.
|
| 59 |
+
|
| 60 |
+
The languages supported by the dataset are:
|
| 61 |
+
- 1. English - `en`
|
| 62 |
+
2. Hindi - `hi`
|
| 63 |
+
3. Bengali - `bn`
|
| 64 |
+
4. Gujarati - `gu`
|
| 65 |
+
5. Urdu - `ur`
|
| 66 |
+
6. Kannada - `kn`
|
| 67 |
+
7. Marathi - `mr`
|
| 68 |
+
8. Punjabi - `pa`
|
| 69 |
+
9. Tamil - `ta`
|
| 70 |
+
10. Telugu `te`
|
| 71 |
+
|
| 72 |
+
The dataset can be used to perform any of the following downstream tasks -
|
| 73 |
+
- 1. Text classification (sentiment analysis models, similarity models)
|
| 74 |
+
2. Text Generation - Fill mask, generation, etc. (samples can be used without the sentiment labels to prepare small corpus for text-generation fine tuning for any of the languages)
|
| 75 |
+
3. Translation - every sentence can be matched with corresponding sentences of the same language code and then used to train translations models between the languages.
|
| 76 |
+
|
| 77 |
+
|
| 78 |
+
We look forward to seeing how you use this dataset!
|
| 79 |
+
|
| 80 |
+
## Author
|
| 81 |
+
Anubhav Singh (for [callchimp.ai](https://callchimp.ai))
|
| 82 |
+
|
| 83 |
+
Contact at: anubhav[at]callchimp.ai for any assistance!
|