IOS Developer turned Machine Learning engineer. I can go from Hardcore to Indie in 0.6 seconds

Rockin' the Suburbs
Joined November 2007
I started working on a NLP task on Kaggle and I've be using Paperspace gradient to train. Once I got my custom docker container done, It makes it really simple to push a change to my repo, then it runs a job on a GCP VM, and saves the model to a folder where I can download
1
0
0
Each job is run on a docker container, which was a little tricky at first, since I need to use an image that contained sklearn and nltk. So I used ufoym/deepo:cpu as the base image and added nltk and joblib
1
0
0
Then, I set up gradient to run as a GitHub action. After each commit, it runs the training job using a config that contains the python command, my docker image(tbass134/deepo-nltk) and the machineType
1
0
0
Tony Dunkin Hung retweeted
I'm really excited to share my latest survey paper on Deep Learning Applications for COVID-19!! šŸ“œ This has been the product of a few months of research, and I really hope it helps people find research projects to impact the pandemic! journalofbigdata.springerope… #100DaysOfMLCode
8
38
131
DALL-E didn't do that great in generating cuban food. Like, where's the moro and cubano?
1
1
0
btw, I'm still in awe on how this is even able to generate images that would be thought to be cuban, or at least its attempt to
0
0
0
I think the hiring process is still broken. Many interviews I've had give a coding test, and if you fail to meet a certain score, you're out. These tests only show how well you can solve a problem, and I don't think having a bad test means you're incapable of doing the job.
1
0
2
A few interviews I've had, the HR person won't even talk to you, until you pass the coding test... I think that's wrong, but I do understand that there are lots of applicants and its impossible to speak to all of us. But still...
0
0
2
Happy New Year Bitches!
0
0
1