How to customize Machine Learning models the simple way

Learn the easiest way to customize pretrained Machine Learning models to your own data.
Speaker: Gus Martins
Watch more:
All Google I/O 2022 Sessions → goo.gle/IO22_A...
ML/AI at I/O 2022 playlist → goo.gle/IO22_M...
All Google I/O 2022 workshops → goo.gle/IO22_W...
Subscribe to TensorFlow → goo.gle/Tensor...
#GoogleIO

[@HumbertoMoura]

Thank you, Gus! Keep it up!

[@osama82405]

hi,from where i can get this notebook?

[@kingalione1171]

these guys are so incompetent of giving any docs on this example. I littlerally had to type the url wtf nice work google

[@subihammed9122]

This colab is not working. All I get is errors and errors and errors

[@subihammed9122]

Its all starting with this error while running the first step:
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
datascience 0.10.6 requires folium==0.2.1, but you have folium 0.8.3 which is incompatible.

[@mominshaikh4084]

@@subihammed9122 !pip install -q tflite-model-maker -U # -U will get you the most recent version or you can specify the version !pip install folium == 0.2.1

[@tarasingh5532]

ਕ

[@kingalione1171]

This example is totally not working. Im trying to fix the little small bugs here and there but the training will still not work due to issues I dont understand:
/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/execute.py in quick_execute(op_name, num_outputs, inputs, attrs, ctx, name)
53 ctx.ensure_initialized()
54 tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,
---> 55 inputs, attrs, num_outputs)
56 except core._NotOkStatusException as e:
57 if name is not None: