yeldarb

40 ms·

391.
by yeldarb 7y ago
Previous discussion: https://news.ycombinator.com/item?id=20301692
392.
by yeldarb 7y ago
I had the pleasure of participating in the synchronous live-stream of the course and loved it. Looking forward to the release of v2 of the library that was prototyped during the course! Roadmap: https://forums.fast.ai/t&#x2F
393.
by yeldarb 7y ago
This doesn't seem particularly newsworthy. It's such a tiny blip that it's immaterial. The RMB & USD have traded in a very tight range throughout 2019.
394.
by yeldarb 7y ago
Looks like a cooler version of the AWS DeepRacer mixed with an old LEGO MindStorm kit.
395.
by yeldarb 7y ago
Part 2 of this course, which will be released later this month, is a step by step walkthrough of building the FastAI and PyTorch libraries from scratch (minus autodiff).
396.
by yeldarb 7y ago
I was in the live version of the most recent course. The last two weeks were guest lectured by Chris Lattner on Swift for TensorFlow. Edit: this appears to be a link to Part 1 which was released in January (and is excellent). The s4tf porti
397.
by yeldarb 7y ago
I thought his manifesto on Medium was pretty thought-provoking. I would identify as a curious-hacker. The bit about the human-blockchain of trust being the best defense against babbling language models made me wonder whether that would forc
398.
A student claims to have replicated the full size version of OpenAI's GPT-2(twitter.com)
4 points by yeldarb 7y ago|1 comments
399.
by yeldarb 7y ago
How would this work? As far as I know neither of them owns the vehicles in their “fleet”. Do they just provide incentives for drivers to buy electric vehicles? Or is the government actually expecting them to put company-owned electric vehic
400.
by yeldarb 7y ago
Maybe not the best choice in light of them deleting this guy’s entire app, db, and backups the other day: https://twitter.com/w3Nicolas/status/1134529316904153089
401.
by yeldarb 7y ago
Came here to ask if they're announcing that they're waiving the $99 fee. This makes it sound like they must be... but I haven't seen any other mention of it. Maybe they jumped the gun a little bit and will announce it on Mond
402.
by yeldarb 7y ago
I did try to rule out question age as a major source of error in the model by training on a sample of ~1 million questions all from ~2012. But it didn’t do any better on those. And (theoretically) duplicates are deleted by the moderators.
403.
by yeldarb 7y ago
I’m not convinced they are randomly distributed but before I started playing with it I was fairly certain they weren’t. Now all I know is I haven’t seen evidence that they aren’t. I could see how there could be some element of luck in wheth
404.
by yeldarb 7y ago
Only title, and text. My original goal was to make a Chrome extension that would give you tips to help you craft a better question. (I added tags to the language model later.)
405.
by yeldarb 7y ago
I’m not convinced. I tried manually predicting high score vs low score as well and couldn’t do so reliably. The language model (viewable at https://stackroboflow.com ) scores much higher accuracy predicting the next word on St
406.
by yeldarb 7y ago
I didn’t try adding in any metadata to the models yet (including time, date, and author). I was just trying to work with the text content of the posts.
407.
by yeldarb 7y ago
Slightly off topic but I’ve been playing with this dataset for a while now to learn ML. It’s remarkably humbling. No NLP approach I’ve tried has been able to predict question score based on content better than my baseline of “choose the mea
408.
by yeldarb 7y ago
I took the recent fastai course that had two lessons about S4TF taught by Chris Lattner and walked away excited. It’s not ready for prime time yet but the roadmap looks great. Building auto-differentiation into the language and having an ea
409.
by yeldarb 7y ago
You mentioned it but, for those who don’t know: As a stopgap while all the data science tooling is being built out for Swift you can use anything from the Python ecosystem (including np and pd) by using the Python interop: let np = Python.i
410.
by yeldarb 7y ago
Can you elaborate on the expected hardware requirements? (Both for fine tuning and inference)
411.
by yeldarb 7y ago
Not sure I understand. The only external dependencies are cdnjs, Firebase, and google fonts.
412.
by yeldarb 7y ago
Aw shoot. Fixing! I "cached" the top questions before deploying and forgot to update the URLs. Edit: fixed!
413.
by yeldarb 7y ago
Hey everyone! Happy to announce the culmination of about a month of digging into ML and NLP. Ask Roboflow is trained on millions of Stack Overflow question/answer pairs to mimic human responses. It's learned lots of interesting th
414.
Show HN: Ask Roboflow, the AI that answers programming questions(ask.roboflow.ai)
10 points by yeldarb 7y ago|7 comments
415.
by yeldarb 7y ago
I'm working on this. I've been training on Stack Overflow and the model has already learned the syntaxes and common coding conventions of a bunch of different languages all on its own. Excited to see what else it's able to do
416.
by yeldarb 8y ago
What challenges are you running into? What sort of data augmentation are you doing?
417.
by yeldarb 8y ago
Interesting, is there a technical reason for not using that extra data to help with the process? For example, another commenter mentioned > Some materials don't really contain any surface details that the algorithms could use to att
418.
by yeldarb 8y ago
I assume this uses GPS data to help with stitching the photos together. Do drones capture orientation data beyond GPS position (for example, pitch, yaw, gimbal direction)? I have tried adding motion graphics to some of my drone videos in Af
419.
by yeldarb 8y ago
http://thisquestiondoesnotexist.com
420.
by yeldarb 8y ago
At least for me, a big part of it has to do with the shift almost entirely to convertible notes. When I made my first angel investment ~10 years ago almost every opportunity was priced. Now almost everything is a note (either with a very hi

More ›