Lina Rojas

Results 1 comments of Lina Rojas

[DialoGPT/data_loader.py](https://github.com/microsoft/DialoGPT/blob/457835e7d8acd08acf7f6f0e980f36fd327ea37c/data_loader.py#L193) Why is **the response** concatenated to the **input_ids** for both the train and validation datasets? Would not this create over-fitted models? Would it be possible to somehow mask the...