I have two scenario. I wanted to know the outcome of these two scenarios!!
- Running full pipeline , Train and Evaluate dataset is different
- Running full pipeline, Train and Evaluate dataset is same
- Can the number of Train documents< Evaluate?
I have two scenario. I wanted to know the outcome of these two scenarios!!
Hi @Ritaman_Baral,
This will give you more confidence on your models accuracy.
Confidence score will not be reliable and not a good practice.
Yes, model will allow it but it will be a poorly trained model will give you poor result.
Thanks,
Ashok ![]()