Fitsum Gaim
commited on
Commit
•
8eb3c17
1
Parent(s):
1f0a04c
Add model card
Browse files
README.md
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: ti
|
| 3 |
+
widget:
|
| 4 |
+
- text: "ዓቕሚ መንእሰይ ኤርትራ [MASK] ተራእዩ"
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
# Pre-trained ELECTRA small for Tigrinya Language
|
| 8 |
+
|
| 9 |
+
We pre-train ELECTRA small on the [TLMD](https://zenodo.org/record/5139094) dataset, with over 40 million tokens.
|
| 10 |
+
|
| 11 |
+
Contained are trained Flax and PyTorch models.
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
## Hyperparameters
|
| 15 |
+
|
| 16 |
+
The hyperparameters corresponding to model sizes mentioned above are as follows:
|
| 17 |
+
|
| 18 |
+
| Model Size | L | AH | HS | FFN | P | Seq |
|
| 19 |
+
|------------|----|----|-----|------|------|------|
|
| 20 |
+
| BASE | 12 | 4 | 256 | 1024 | 14M | 512 |
|
| 21 |
+
|
| 22 |
+
(L = number of layers; AH = number of attention heads; HS = hidden size; FFN = feedforward network dimension; P = number of parameters; Seq = maximum sequence length.)
|