jkraybill

I had this error; I was using a non-GPU build of Tensorflow. Make sure you install tensorflow-gpu or use a build with GPU support.

@GNU-Linuxer - are you using a n1-highcpu-96 instance? How long did fine-tuning take under CPU?

@GNU-Linuxer thanks for the follow-up! I'm going to try the same setup.

Just a follow-up, I was able to successfully do full-depth training on the full-sized model, using Adam, on an Amazon r4.4xlarge EC2 instance (16 vCPU, 122GB RAM). CPU training is...

@tomerwul per my last post I was using batch size of 1 so can't speak to when that tops out, on a r4.4xlarge instance which is CPU only. It's obviously...

Just FYI, I have been able to fine-tune a sub-set of variables under Colab with good results. My forked notebook (via Tenoke's fork) demonstrating how to do this is here:...

@saippuakauppias FWIW I just learned today that Linode is running a paid pilot with 24GB Quadro RTX's. Haven't used them personally.

@abvgdeabvgde2 thanks so much for that -- I was pulling my hair out trying to install python 3.10 and all dependencies in Colab, this was a nice elegant fix! 100%...