Hello Giri,
I have to admit, it is bit difficult to give any advice without better knowledge of your model.
Generally speaking, the Count function returns the number of unique values of an attribute within the given context.
How did you upload this Attribute, please? Is this your ConnectionPoint/PrimaryKey or is there any other ConnectionPoint/PrimaryKey set for related dataset?
I believe, this is what happened here - since these values are not unique, only the one of them is loaded, deduplicated during the load (in other words, although there is probably a difference in other fields of your CSV - it is still considered as the same and only the latest is loaded).