Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,30 @@
|
|
| 1 |
---
|
| 2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
+
datasets:
|
| 4 |
+
- oscar
|
| 5 |
+
language:
|
| 6 |
+
- uk
|
| 7 |
+
library_name: transformers
|
| 8 |
+
pipeline_tag: text-generation
|
| 9 |
---
|
| 10 |
+
|
| 11 |
+
# GPT2 Ukrainian
|
| 12 |
+
|
| 13 |
+
A generative language model for the Ukrainian language follows the [GPT-2 architecture](https://huggingface.co/gpt2) (124M parameters).
|
| 14 |
+
|
| 15 |
+
- hidden size: 768
|
| 16 |
+
- number of heads: 12
|
| 17 |
+
- number of layers: 12
|
| 18 |
+
- seq length: 1024
|
| 19 |
+
- tokens: 11238113280 (3 epochs)
|
| 20 |
+
- steps: 57167
|
| 21 |
+
|
| 22 |
+
## Training data
|
| 23 |
+
|
| 24 |
+
- OSCAR
|
| 25 |
+
- Wikimedia dumps
|
| 26 |
+
|
| 27 |
+
## License
|
| 28 |
+
|
| 29 |
+
MIT
|
| 30 |
+
|