Commit
·
6fdce9f
1
Parent(s):
cb1bfd0
Update README.md
Browse files
README.md
CHANGED
@@ -4,22 +4,25 @@ license: cc-by-nc-4.0
|
|
4 |
language:
|
5 |
- en
|
6 |
---
|
7 |
-
# Jellyfish-
|
8 |
<!-- Provide a quick summary of what the model is/does. -->
|
9 |
<img src="https://i.imgur.com/d8Bl04i.png" alt="PicToModel" width="330"/>
|
10 |
|
11 |
## Model Details
|
12 |
-
Jellyfish is a model designed specifically for data preprocessing tasks,
|
13 |
|
14 |
-
We fine-tuned [Open-Orca/OpenOrca-Platypus2-13B](https://huggingface.co/Open-Orca/OpenOrca-Platypus2-13B) using data
|
|
|
15 |
|
16 |
-
|
|
|
|
|
|
|
17 |
|
18 |
-
|
19 |
-
|
20 |
-
As the names suggest, the main version focuses on providing accurate, direct answers. In contrast, Jellyfish-reasoning can generate not only the final answer to data processing queries but also the reasons behind its decisions.
|
21 |
|
22 |
- **Developed by:** Haochen Zhang, Yuyang Dong, Chuan Xiao, Masafumi Oyamada
|
|
|
23 |
- **Funded by:** NEC Corporation, Osaka University
|
24 |
- **Language(s) (NLP):** English
|
25 |
- **License:** Non-Commercial Creative Commons license (CC BY-NC-4.0)
|
|
|
4 |
language:
|
5 |
- en
|
6 |
---
|
7 |
+
# Jellyfish-13B
|
8 |
<!-- Provide a quick summary of what the model is/does. -->
|
9 |
<img src="https://i.imgur.com/d8Bl04i.png" alt="PicToModel" width="330"/>
|
10 |
|
11 |
## Model Details
|
12 |
+
Jellyfish-13B is a 13B large language model designed specifically for data managment and preprocessing tasks, such as entity matching, data imputation, error detection, and schema matching.
|
13 |
|
14 |
+
We fine-tuned [Open-Orca/OpenOrca-Platypus2-13B](https://huggingface.co/Open-Orca/OpenOrca-Platypus2-13B) using data preprocessing tasks datasets.
|
15 |
+
Its performance is competitive, standing up well against prior state-of-the-art algorithms and OpenAI GPT 3.5 and GPT 4, which is only 13B model and run locally for low cost and data security.
|
16 |
|
17 |
+
We have released two versions of Jellyfish: the Jellyfish-13B and Jellyfish-13B-Reasoning.
|
18 |
+
As the names suggest, Jellyfish-13B focuses on providing accurate, direct answers.
|
19 |
+
In contrast, Jellyfish-13B-Reasoning is distilled by GPT4, which is fine-tuned with data of the reasons and chain-of-thought responses for solving data preprocessing tasks
|
20 |
+
generated by OpenAI GPT-4.
|
21 |
|
22 |
+
Jellyfish paper will coming soon!
|
|
|
|
|
23 |
|
24 |
- **Developed by:** Haochen Zhang, Yuyang Dong, Chuan Xiao, Masafumi Oyamada
|
25 |
+
- **Contact: [email protected]**
|
26 |
- **Funded by:** NEC Corporation, Osaka University
|
27 |
- **Language(s) (NLP):** English
|
28 |
- **License:** Non-Commercial Creative Commons license (CC BY-NC-4.0)
|