Shipeng Wang

Hi, In your paper, you said the "next-utterance classification" task and "language modeling" task were trained in a multi-task learning setting, and also in train.py, there is a function load...

如题，打算引入 trl 库的KTOTrainer吗？

Inquiry about the impact of gradient checkpointing on KL divergence estimation.

1

I am currently working on experiments of DPO and KTO Trainer on private dataset. I am considering using gradient checkpointing to reduce memory usage during backpropagation, but I am unsure...

qlora taining on qwen1.5-15b-chat

3

训练qwen1.5-14b-chat，遇到下面的报错，transformers==4.38.2 RuntimeError( "Unsloth: Tokenizer's pad_token cannot be = eos_token, and we couldn't find a\n"\ "replacement of either