XuanRen4470

我跑dpo的时候遇到了ZeroDivisionError: integer division or modulo by zero的error

1

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction ``` Loading checkpoint shards: 0%| | 0/3 [00:00

究竟怎么做dpo呀

6

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction 我已经严格按照example里dpo的流程做了(lora) 1. train sft model 2. 把sft adapter的路径给adapter_name_or_path， 把mistral 7b 的路径（没merge）给model_name_or_path 3. 推理时 把mistral...

how to evaluate the model?

while I am doing the second last step Merge Reward adapter into Model: python merge_peft_adapter.py --model_name ./reward_model_vicuna-7b I got the following error ``` Traceback (most recent call last): File "/home/xuan/anaconda3/envs/vicuna_lora/lib/python3.9/site-packages/peft/tuners/lora.py",...

Unable to download the dataset from huggingface

2

I tried to use huggingface to download the dataset ``` from datasets import load_dataset dataset = load_dataset("liminghao1630/API-Bank") ``` but it gave me the error ``` An error occurred while generating...