ARIELDENG comments

Results 6 comments of


                                            ARIELDENG

使用SFT后的FP32模型进行生成，报错RuntimeError: where expected condition to be a boolean tensor, but got a tensor with dtype Half

通过zero_to_fp32.py文件将上述多组pt文件转成pytorch_model.bin，此外在index.json里将所有参数都指向了pytorch_model.bin

ModuleNotFoundError: No module named 'transformers_modules.local.custom_autotune'

the same

微调运行run.sh报错

> > 按照步骤复制之后微调报错 ValueError: The port number of the rendezvous endpoint 'None:None' must be an integer between 0 and 65536. 请问是哪里出问题了吗？ > > 检查一下 `configs/sft.yaml` 文件中的 `main_process_ip` 和 `main_process_port`是否被正确指定请问您这两个参数您是怎么设置的呢？

微调运行run.sh报错

> 我把configs/sft.yaml 文件中的 main_process_ip 和 main_process_port的俩个null都改成0了，程序就不报这个错了，但是会报timeout。请问这个问题您是最后这么解决的还有timeout的话可以把num_machines改成1哈，应该是通信等待时间过长timeout捏

format for inference in code completion

thanks for your attention, but the thing is that the output won't stop when I apply this formatting, just like you @xcxhy ![image](https://github.com/bigcode-project/starcoder2/assets/32135928/a2569ba6-e05a-434e-9cae-216349411716) However, it seems to be following a...

format for inference in code completion

> yes `` is the token we use to separate files so you can use it as a stop token. The `` token was used to separate repositories since we...