HF RL Explorer

Feature Request: Russian Language Support for SQuAD Model with Training Improvements

Feature Request: Russian Language Support for SQuAD Model with Training Improvements: a task in LegoFlow-SWE (Harbor dataset). The model should follow the same external interface as the current English SQuAD model – accepting a context string and a question string, and returning a predicted answer…

The task

The model should follow the **same external interface** as the current English SQuAD model – accepting a context string and a question string, and returning a predicted answer string along with its start position in the original context.

Part of Lego-X/LegoFlow-SWE.