Commit ·
57fe380
1
Parent(s): 1647663
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
library_name: peft
|
| 3 |
license: apache-2.0
|
|
|
|
| 1 |
+
|
| 2 |
+
This is a Finetuning of GPT-J-6B using LoRa - https://huggingface.co/EleutherAI/gpt-j-6B
|
| 3 |
+
|
| 4 |
+
The dataset is the cleaned version of the Alpaca dataset - https://github.com/gururise/AlpacaDataCleaned
|
| 5 |
+
|
| 6 |
+
A model similar to this has been talked about
|
| 7 |
+
|
| 8 |
+
The performance is good but not as good as the orginal Alpaca trained from a base model of LLaMa
|
| 9 |
+
|
| 10 |
+
This is mostly due to the LLaMa 7B model being pretrained on 1T tokens and GPT-J-6B being trained on 300-400M tokens
|
| 11 |
+
|
| 12 |
+
You will need a 3090 or A100 to run it, unfortunately this current version won't work on a T4.
|
| 13 |
+
|
| 14 |
+
|
| 15 |
---
|
| 16 |
library_name: peft
|
| 17 |
license: apache-2.0
|