Spaces:
Running
Running
Add model and dataset links to README
Browse files
README.md
CHANGED
@@ -92,6 +92,13 @@ The model weights and SentencePiece model are stored on Hugging Face Hub and are
|
|
92 |
4. **Text Generation**: At inference time, nucleus sampling is used to select the next word with a balance of creativity and coherence.
|
93 |
5. **Formatting**: Generated text is automatically formatted into lines with alternating indentation for aesthetic presentation.
|
94 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
95 |
## Limitations
|
96 |
|
97 |
- The current model was trained on a relatively small dataset (~1300 lines), which may occasionally result in repetitive patterns.
|
@@ -111,4 +118,3 @@ This project is licensed under the Apache License 2.0 - see the [LICENSE](LICENS
|
|
111 |
## Acknowledgements
|
112 |
|
113 |
- Poetry is the rhythmical creation of beauty in words - Edgar Allan Poe
|
114 |
-
|
|
|
92 |
4. **Text Generation**: At inference time, nucleus sampling is used to select the next word with a balance of creativity and coherence.
|
93 |
5. **Formatting**: Generated text is automatically formatted into lines with alternating indentation for aesthetic presentation.
|
94 |
|
95 |
+
## Model and Dataset
|
96 |
+
|
97 |
+
- **Model**: You can find the complete model, weights, and training notebooks on Hugging Face:
|
98 |
+
[Mehfil-e-Sukhan on Hugging Face](https://huggingface.co/zaiffi/Mehfil-e-Sukhan)
|
99 |
+
- **Dataset**: The model was trained on the Roman Urdu Poetry dataset available on Kaggle:
|
100 |
+
[Roman Urdu Poetry Dataset](https://www.kaggle.com/datasets/mianahmadhasan/roman-urdu-poetry-csv)
|
101 |
+
|
102 |
## Limitations
|
103 |
|
104 |
- The current model was trained on a relatively small dataset (~1300 lines), which may occasionally result in repetitive patterns.
|
|
|
118 |
## Acknowledgements
|
119 |
|
120 |
- Poetry is the rhythmical creation of beauty in words - Edgar Allan Poe
|
|