@ltirrell Sure. I was stuck at lesson 1 for a while (mostly distracted by a lot content available on Internet )
I will start lesson 2 next week.
I am also using Google colab and found it straightforward.
@ltirrell Sure. I was stuck at lesson 1 for a while (mostly distracted by a lot content available on Internet )
I will start lesson 2 next week.
I am also using Google colab and found it straightforward.
How do I get this all working in Spyder? I canāt figure it out. Canāt even import fastai without an error
IDK much but would suggest using colab or any other service .
Use the anaconda python distribution
can we guys form a group on signal or discord ?
I built a traffic classifier that would take CCTV images and decide if itās heavy_traffic or not. Not sure if I am doing it right but I got 6% error rate using Resnet50 which seems pretty good.
I have some screenshots at my Twitter: https://twitter.com/jjsee/status/1166972223732584448
It was actually quite straight forward to build this as I only have to modify lesson-2 slightly to fit my classes.
I used pandas for reading .json files in the PASCAL dataset and it is throwing an error āArrays must be all same lengthā.Does anyone have any suggestions to correct the error or any other way to read the .json files. Stack Overflow is not helpingā¦
I started following you on twitter and like Jeremy said itās really really hard to overfit a deep learning model so you are in a good spot with that 6 percent error, good work
I am in love with google colab. Free + connectivity with google drive+ easy setup+good community in fastai to help with colab. <3
This is such a good idea. Have you guys already formed one?
I like Sypder alot but no no for DataScience
Iām getting this same error. Did you find a solution? Iāve tried reinstalling pytorch, but no luck.
Disregard. I found my problem, didnāt have an Nvidia driver installed.