How to use replace_na if column is integer

Question

I am trying to replace "NA" in a dataframe that has 3 columns: Day, Month, and Temperature.

nas <- mydata %>% replace_na(list(Temp = "unknown"))

but I get this error: Error in vec_assign(): ! Can't convert replace$Temp to match type of data$Temp .

Example of two records in my data

 Day    Month   Temp
 30     7       68
 31     7       NA

Thanks!