I have followed the instrucions in data_collection notebook an it looks ok. At the end, I am creating a dataset.zip file. Now I am asked to download the zip-file. Using the webbrowser I am working on the robot computer Notebooks/collision_avoidance directory so I do not understand this. Download to where?
Next I open the train_model notebook. Now I am asked to upload and extract the dataset.zip file. Uplode from where to where? Since I am still working on the same directory on the robot, this makes no sence to me. Please explain how this is supposed to work. Extracting the dataset.zip file will taget the existing dataset directory, generating a warning.
After running the data_collection notbook I have a dataset folder with all the images split in two folders blocked and free.
With the zip command I zip the dataset folder and create a zip file in the collision_avoidance folder
The Download command seams to create a file named “Obekräftade 154904.crdownload” in the Downloaded files directory on my swedish windows computer.
What confuses me is that now I am asked to upload a file which I guess is the file I just dowloaded.
Since I will use the train_model notebook residing in the collision_avoidance folder I will upload it to the same folder from where I downloaded it, but now it has the strange name that it got from the download.
The unzip command in the train_model notbook will however unzip the original dataset.zip file to the already existing dataset folder which will insert this line in my notebook:
replace dataset/blocked/2e1e294c-a5b1-11ef-9b4f-16f63a1aa8c9.jpg? [y]es, [n]o, [A]ll, [N]one, [r]ename:
To me the zip, download, upload and unzip seams redundant.
I am a bit confused and don’t know how to proceed.
Yes, the zip-file is saved in the notebook folder, but when downloaded, it is downloaded to my “Hämtade filer” folder on my Windows 10 desktop and renamed to “Obekräftade 471827.crdownload”.
The notebook instructions now says:
" Next, we’ll need to upload this data to our GPU desktop or cloud machine (we refer to this as the host ) to train the collision avoidance neural network. We’ll assume that you’ve set up your training machine as described in the JetBot WiKi. If you have, you can navigate to http://<host_ip_address>:8888 to open up the Jupyter Lab environment running on the host. The notebook you’ll need to open there is called collision_avoidance/train_model.ipynb".
I don’t understand this. As I understand, my Jetbot is the host, where I collected my data, so I don’t understand why I shall zip, download, upload and unzip the dataset since eveything resides in the same Notebooks/collision_avoidance directory which now contains this directory and these files:
dataset
best_model.pth
data_collection.ipynb
dataset.zip
live_demo.ipynb
train_model.ipynb
When I open the train_model.ipynb I am asked to upload the dataset.zip file, but why, it is already there.
Then I am asked to unzip the dataset.zip file which will try to overwrite the dataset directory and will output the following message to my notebook:
replace dataset/blocked/2e1e294c-a5b1-11ef-9b4f-16f63a1aa8c9.jpg? [y]es, [n]o, [A]ll, [N]one, [r]ename:
Next it says:
You should see a folder named dataset appear in the file browser.
but this folder was alredy created when I ran the data_collection.ipynb notebook.
Have I totally missunderstood this or is there something wrong with the instructions?
I have the SparkFun Kit refered to in the heading of this post. The instructions I have followed are in the Notbooks preinstalled on the SD card that comes with the kit.
What do you mean by asking another forum? Isn`t this the SparkFun forum where I can ask questions regarding Kit-18486?
We just sell the kit and show how to assemble the pieces; the guides for jetbots are all from outside sources (We use Nvidia’s, but there are others)…we do not have any control over what the example instructions say. Feel free to file an issue on github, although I believe whatever misunderstanding that is happening here is likely not preventing the project from moving forward…it’s ok to try multiple methods, duplicate files, backup copies, re-collected data, etc - if the file is already there, great
You must however take full responsibility for the SD-card and what is on it since it is included in the kit.
You refer to github. When I looked there, I found a JetBot repository and it seams to match what’s on the SD-card. Most of it was last updated between 2 and 4 years ago, so it is completely out of date.
As a newbie like me, it is important to have correct information, to be able to learn how to use the kit. I must say I am a bit disapointed. The quality of the kit does not meet my expectations. When I buy a kit like this, where software is included I expect also the software to be up to date. I can run the robot using the software, but the AI part is not much help and needs updating. The current version of JetPak is 6.1 and I think the version included on the SD-card is 4 something.
I think you should have someone update the software and offer a new image to customers that they can flash to the sd-card to get full use of your product.