Open nxfi777 opened 1 month ago
Thanks for pointing this out! I understand the current implementation relies on an embedding layer, which might not be ideal for floating-point data. I'd love to assign this task to you - would you be willing to take on the task of modifying the implementation to accommodate floating-point data? If so, I can provide the necessary guidance and support. Let me know if you're interested and we can discuss further!
Current implementation seems to require an embedding layer. This is inconvenient for models training on floating-point data.