I want to search for my own images of people with different professions, and upload them. The notebook’s code is set up just to upload the images of bears. How can I replace the code in the nb to load my own image urls that are in my txt files on my computer?
Anyone know how I can reach people who will be able to answer these types of questions?
You need to add it to whatever file system you are using for your jupyter notebook, there should be a way to upload it. Then set your
path to lead to where your csv’s are located. Then do the same thing as seen in the lesson 2 notebook, call
download_images where your path/file is set to your new csv document, your destination is set to your location you want to store to, and your maximum pictures you want downloaded. You can ensure if it is going to the right location beforehand by just running a cell with
path/file to see what it outputs. If it doesn’t go to the right place, adjust. Hope this helps.
how do i set my path to lead where my files are located on my computer?
why does it label them .csv? shouldn’t it be .txt?
First do a path = Path(‘’) and path.ls() this will show you where the current working directory you’re looking at is. Then you can move the files from there. A CSV document has a comma separated delimiter, so each line instead of a new line, is separated by a comma. This is used often in datasets.
Wow! You are a very big help! Tell your boss I said that. You should get a raise.
i typed into the nb
path = Path('') path.ls()
and got a reponse that
name 'Path' is not defined
Oh, I see my error. I hadn’t yet loaded the top cell of import fast ai. Any simple way to run all cells of a nb from top to bottom?
ANyhow, now when I type into the nb
path = Path('') path.ls()
I get a response, as follows, which I don’t know what it means, nor what am I supposed to do with it…
I think I accomplished #1, but I don’t know how to do #2. Please help? Should I upload a screen shot?
sry i need someone to hold my hand thru this.
I think you can just write
path='../working/model/' — since you are working on kaggle kernels.
But, this path is where fastai will store all the model related details.
For reading image files, use ImageDataBunch. Similarly, you can put your images in the
'../input/files’ folder and read from there.
It is hard for me to know what you mean without seeing it right in the nb.
Do you mean that instead of what the nb says, i.e.
path = Path('data/bears')
I should replace that with
or as an equally acceptable alternative,
Also I still don’t understand how to upload the files from my computer – they are on my desktop.