Yi Dong
Yi Dong
Could you try it locally in your container instead of Colab? I don't think Colab has enough compute resource to train a GPT.
you need to use `megatron_t5_prompt_learning.py` to run the conversion. Currently only `GPT` and `BERT` models are supported. For `T5`, we don't have the implementation yet.
@harrism. The implementation I have is just adding a weight term to the time series items in the rolling window fun. So it should be straight forward implementation on top...
We have seen some folks at FSI who are interested in the official EWM function. Check this issue we got from gQuant project. I can fix it in a hacky...
> @yidong72 Shall we contact someone to conduct the CI? I think we need to add "RUN CICD" label to it. let's see if it kicks off the tests