Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,34 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
datasets:
|
4 |
+
- tatsu-lab/alpaca
|
5 |
+
language:
|
6 |
+
- en
|
7 |
---
|
8 |
+
# Model Card for Flan-Alpaca-GPT4-base-3k
|
9 |
+
|
10 |
+
This model was obtained by fine-tuning the `google/flan-t5-base` model on the tatsu-lab/alpaca dataset with the max_source_length option set to 3048. The instructions at the following repository were used for fine-tuning: https://github.com/declare-lab/flan-alpaca
|
11 |
+
The goal of this model was a learning exercise to determine if setting a higher max_source_length resulted in the model interpreting larger prompts during inference.
|
12 |
+
|
13 |
+
### Model Description
|
14 |
+
|
15 |
+
- **Language(s) (NLP)**: English
|
16 |
+
- **Finetuned from model:** google/flan-t5-base
|
17 |
+
|
18 |
+
## How to use
|
19 |
+
|
20 |
+
```python
|
21 |
+
from transformers import pipeline
|
22 |
+
|
23 |
+
prompt = "Write an email about an alpaca that likes flan"
|
24 |
+
model = pipeline(model="evolveon/flan-alpaca-gpt4-base-3k")
|
25 |
+
model(prompt, max_length=3048, do_sample=True)
|
26 |
+
|
27 |
+
# Dear AlpacaFriend,
|
28 |
+
# My name is Alpaca and I'm 10 years old.
|
29 |
+
# I'm excited to announce that I'm a big fan of flan!
|
30 |
+
# We like to eat it as a snack and I believe that it can help with our overall growth.
|
31 |
+
# I'd love to hear your feedback on this idea.
|
32 |
+
# Have a great day!
|
33 |
+
# Best, AL Paca
|
34 |
+
```
|