I tried training a language model with fastai on Modal.
First I attempted it in a standalone Modal script.
I first wrote a script to unpack the data to a volume, then ran the fit_one_cycle function with the learner.
I ran into an issue with counter.pkl sort of similar to this issue but I haven’t figured out how to resolve it yet.
On a whim, I checked to see if I could run a Jupyter notebook on Modal. Apparently, you can!
I ran
modal launch jupyter --gpu a10g
then started to mess around a bit. I uploaded the imdb data again then untarred it.
I was noticing occasionally laggy/busy behavior at times that I wasn’t expecting. This experiment ended for the day when I tried to run
learn = language_model_learner(
dls_lm, AWD_LSTM, drop_mult=0.3, metrics=[accuracy, Perplexity()]
).to_fp16()
I got a SSLCertVerificationError and wasn’t prepared to spend more time debugging.