Difference between encoder_only and decoder_only?

[ccdv-ai]

What is the difference between the encoder_only script and the decoder_only script if we use last_token as pooling method and m3_kd_loss as loss? Can lora be used with encoder only?

Edit: for lora with encoder m3: https://github.com/ccdv-ai/FlagEmbedding not much time to work on it