Update README.md
Browse files
README.md
CHANGED
|
@@ -35,7 +35,19 @@ The tool can be found [here](https://github.com/ayaka14732/lihkg-scraper).
|
|
| 35 |
Please also check out the [Bart model](https://huggingface.co/Ayaka/bart-base-cantonese) created by her.
|
| 36 |
|
| 37 |
|
| 38 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
Please refer to the [script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling)
|
| 41 |
provided by Huggingface.
|
|
@@ -44,8 +56,6 @@ provided by Huggingface.
|
|
| 44 |
The model was trained for 400,000 steps with batch size 5 (~2epoches) on 2 NVIDIA Quadro RTX6000 for around 40 hours at the Research Computing Services of Imperial College London.
|
| 45 |
|
| 46 |
|
| 47 |
-
|
| 48 |
-
|
| 49 |
### How to use it?
|
| 50 |
```
|
| 51 |
from transformers import AutoTokenizer
|
|
@@ -62,6 +72,7 @@ string = output[0]['generated_text'].replace(' ', '')
|
|
| 62 |
print(string)
|
| 63 |
```
|
| 64 |
|
|
|
|
| 65 |
### Framework versions
|
| 66 |
|
| 67 |
- Transformers 4.26.0.dev0
|
|
|
|
| 35 |
Please also check out the [Bart model](https://huggingface.co/Ayaka/bart-base-cantonese) created by her.
|
| 36 |
|
| 37 |
|
| 38 |
+
|
| 39 |
+
### Limitations
|
| 40 |
+
The model was trained on ~10GB of data scrapped from LIHKG.
|
| 41 |
+
It might contain violent and rude languages so as the text generated by the model.
|
| 42 |
+
Please do not use it for anything other than research or entertainment.
|
| 43 |
+
|
| 44 |
+
|
| 45 |
+
The comments on LIHKG also tend to be very short.
|
| 46 |
+
Thus the model cannot generate anything more than a line. In a lot of occasions might not even generate new tokens.
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
|
| 50 |
+
### Training procedure
|
| 51 |
|
| 52 |
Please refer to the [script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling)
|
| 53 |
provided by Huggingface.
|
|
|
|
| 56 |
The model was trained for 400,000 steps with batch size 5 (~2epoches) on 2 NVIDIA Quadro RTX6000 for around 40 hours at the Research Computing Services of Imperial College London.
|
| 57 |
|
| 58 |
|
|
|
|
|
|
|
| 59 |
### How to use it?
|
| 60 |
```
|
| 61 |
from transformers import AutoTokenizer
|
|
|
|
| 72 |
print(string)
|
| 73 |
```
|
| 74 |
|
| 75 |
+
|
| 76 |
### Framework versions
|
| 77 |
|
| 78 |
- Transformers 4.26.0.dev0
|