Update README.md
Browse files
README.md
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
---
|
| 2 |
language:
|
| 3 |
- ar
|
| 4 |
-
---
|
|
|
|
|
|
| 1 |
---
|
| 2 |
language:
|
| 3 |
- ar
|
| 4 |
+
---
|
| 5 |
+
The QARiB-Sentiment-Reviews-Tweets (QSRT) Model is the continuation of training from the last checkpoint of the QSR model for an additional 300k training steps (600k in total with QSR training steps). The QSRT model was trained for roughly 52 epochs using sentiment dataset contains both tweets and reviews (consists of 1,433,657 sentences).
|