Kexin Wang

Hi @kingafy, GPL needs only a corpus.jsonl file (data sample is [here](https://github.com/UKPLab/gpl/blob/main/sample-data/generated/fiqa/corpus.jsonl)) for minimal running. Specifically, you need three steps: 1. Prepare your corpus in the same format as this...

Hi @christopherfeld, could you please try this toy example: https://github.com/UKPLab/gpl/issues/5#issuecomment-1144256494 and let me know whether it is runnable in your case?

Hi @christopherfeld, I have created a google colab notebook showing the running results of the toy example: https://colab.research.google.com/drive/1Wis4WugIvpnSAc7F7HGBkB38lGvNHTtX?usp=sharing Please have a look:). BTW, I think the error might be due...

Hi @christopherfeld, sorry for the delay. I will look into this in the near future. Your idea sounds great and I think we can have this option during pip install

Hi, It seems that the latest version of HF's trainer will only create a scaler when enabling `sharded_ddp` https://github.com/huggingface/transformers/blob/fa6107c97edf7cf725305a34735a57875b67d85e/src/transformers/trainer.py#L637 Does this influence the `tevatron` code? Thanks

I think now HF's trainer uses the accelerate's scaler https://github.com/huggingface/transformers/issues/25021#issuecomment-1647349987

@Rachneet a file is missing: https://github.com/UKP-SQuARE/square-core/blob/49ef708a2193b9057a9b61a12a6b65b63ebe3935/docker-compose.ytt.min.yaml#L218

Does anybody know how to add CI tests for this minimal version? @timbmg @Rachneet We can then make it easier for maintaining

Hi @lijiaoyang, I have re-implemented BERT-flow under the Pytorch framework. I carefully transferred each line of the original TF code into the Pytorch one and the intermediate results are equal....

Hi @Muennighoff, Yeah, we tried that. Actually what you said seems to be exactly `SBERT-base-nli-v2`, `SBERT-base-nli-stsb-v2` (zero-shot models) and `SBERT-supervised` (in-domain supervised) in Table 2. All of them were trained...