masoud-nasiripour
commited on
Commit
•
b051839
1
Parent(s):
95253a3
Create README.md
Browse filesThis model get date in 4 different informal format and convert it to a formal date.
1. month in Persian, year and day in English numbers. e.g. "1 فروردین 1400"
2. month and day in Persian format, year in English numbers. e.g. "یکم فروردین 1400"
3. all in Persian format. e.g. "یکم فروردین سال هزار و چهارصد"
4. month and year in Persian, day in English numbers. e.g. "1 فروردین سال 1400"
we create a dataset for these types and train the model on about 80,000 pair of (informal date, formal date)
We used Transformer Encoder model with 6 layer alongside with persian-llama2 tokenizer.