And change the instructions / workflow to simply download it. That way we eliminate the need to use Python at all, and things become more robust. One would only use the Python create_model.py script if one wants to re-generate the model.dat.
We should upload all 4 versions, so probably something like:
model-fastGPT-124M.dat
model-fastGPT-1558M.dat
Not sure if we should put the version number into the filename itself as well or not.
We should only do this after #30 is fixed, to prevent downloading incompatible model.dat (from older or newer versions of fastGPT).
And change the instructions / workflow to simply download it. That way we eliminate the need to use Python at all, and things become more robust. One would only use the Python
create_model.py
script if one wants to re-generate themodel.dat
.We should upload all 4 versions, so probably something like:
model-fastGPT-124M.dat
model-fastGPT-1558M.dat
Not sure if we should put the version number into the filename itself as well or not.
We should only do this after #30 is fixed, to prevent downloading incompatible
model.dat
(from older or newer versions of fastGPT).