KaedinLian

Results 3 issues of KaedinLian

**The bug** get .��� character after running guidance writen by chinese in jupyter nokebook. And in normal .py file, it works fine. **To Reproduce** Give a full working code snippet...

When I click regenerate button, I got ***NameError: name 'history' is not defined*** Is it normal to not define the history variable in advance? How can I solve my problem?...

enhancement

## training_commands: ```bash openrlhf.cli.train_dpo --save_path ./checkpoint/llama3-8b-dpo-lora --save_steps -1 --logging_steps 1 --eval_steps -1 --train_batch_size 24 --micro_train_batch_size 1 --pretrain unsloth/llama-3-8b-Instruct --bf16 --max_epochs 1 --max_len 2048 --zero_stage 3 --learning_rate 5e-6 --beta 0.1 --dataset...