Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,38 +1,12 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
dtype: string
|
| 14 |
-
- name: user_id
|
| 15 |
-
dtype: string
|
| 16 |
-
- name: messages
|
| 17 |
-
list:
|
| 18 |
-
- name: content
|
| 19 |
-
dtype: string
|
| 20 |
-
- name: role
|
| 21 |
-
dtype: string
|
| 22 |
-
splits:
|
| 23 |
-
- name: train
|
| 24 |
-
num_bytes: 492020833
|
| 25 |
-
num_examples: 202362
|
| 26 |
-
- name: test
|
| 27 |
-
num_bytes: 3410666
|
| 28 |
-
num_examples: 1750
|
| 29 |
-
download_size: 263068673
|
| 30 |
-
dataset_size: 495431499
|
| 31 |
-
configs:
|
| 32 |
-
- config_name: default
|
| 33 |
-
data_files:
|
| 34 |
-
- split: train
|
| 35 |
-
path: data/train-*
|
| 36 |
-
- split: test
|
| 37 |
-
path: data/test-*
|
| 38 |
-
---
|
|
|
|
| 1 |
+
This is a converted version of the Aya dataset into Tulu SFT training format.
|
| 2 |
+
|
| 3 |
+
The conversion script can be found in our [open-instruct](https://github.com/allenai/open-instruct/blob/main/scripts/data/sft/aya.py) repo.
|
| 4 |
+
The conversion took the following parameters:
|
| 5 |
+
- apply_keyword_filters: False
|
| 6 |
+
- apply_empty_message_filters: False
|
| 7 |
+
- push_to_hub: True
|
| 8 |
+
- hf_entity: VGraf
|
| 9 |
+
- converted_dataset_name: aya_dataset_converted
|
| 10 |
+
- local_save_dir: None
|
| 11 |
+
|
| 12 |
+
Please refer to the [original dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) for more information about this dataset and the license.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|