Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,27 @@
|
|
| 1 |
---
|
| 2 |
-
license:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
license: llama2
|
| 3 |
+
task_categories:
|
| 4 |
+
- conversational
|
| 5 |
+
language:
|
| 6 |
+
- ja
|
| 7 |
+
size_categories:
|
| 8 |
+
- 1K<n<10K
|
| 9 |
---
|
| 10 |
+
This dataset is based on the Japanese version of Wikipedia dataset and converted into a multi-turn conversation format using llama2Pro8B. After generating 10,000 conversations and screening, only about 3,000 were usable, so I will publish them in this state first.
|
| 11 |
+
Since it is a llama2 license, it can be used commercially for services with less than 700 million monthly users.
|
| 12 |
+
|
| 13 |
+
Some strange dialogue may be included as it has not been screened by humans.
|
| 14 |
+
|
| 15 |
+
We generated 10,000 conversations over 24 hours on an A100 80GBx7 machine and automatically screened them.
|
| 16 |
+
|
| 17 |
+
# Model
|
| 18 |
+
https://huggingface.co/spaces/TencentARC/LLaMA-Pro-8B-Instruct-Chat
|
| 19 |
+
|
| 20 |
+
# Dataset
|
| 21 |
+
https://huggingface.co/datasets/izumi-lab/wikipedia-ja-20230720
|
| 22 |
+
|
| 23 |
+
# Compute by
|
| 24 |
+
Tsuginosuke AI SuperComputer
|
| 25 |
+
FreeAI Ltd.
|
| 26 |
+
|
| 27 |
+
https://free-ai.ltd
|