neutralino1

9 ms·

1.
by neutralino1 8mo ago
Asking people to change their ways is pointless. When something is systemic, only a systemic solution can work. I have become intimately convinced that engagement-based feeds are the root of many evils of our time, loneliness included. Here
2.
by neutralino1 1y ago
You win today.
3.
by neutralino1 1y ago
It seems to me Rails has been doing this but better for years. It definitely keeps atomic and historical migrations, but also maintains a schema.sql file that can be loaded as a one-off (e.g. for mock DBs in tests).
4.
How OpenAI's o1 changes the LLM training picture(airtrain.ai)
8 points by neutralino1 2y ago|0 comments
5.
Show HN: Fineweb-Edu-Fortified dataset: Fineweb-Edu deduped, embeddings included(huggingface.co)
22 points by neutralino1 2y ago|0 comments
6.
by neutralino1 2y ago
Yeah sorry about that, this is a data dashboard. Not optimized for mobile viewing.
7.
by neutralino1 2y ago
We debated doing 2D vs 3D and 3D brought a bunch of usability issues. We also noticed most SOTA embedding visualizations were 2D and already yielded good insights.
8.
by neutralino1 2y ago
The original dataset is located at [1] (not our HF account). HN data is directly available via the HN API [2]. The privacy policy you point to does not cover HN posts. [1] https://huggingface.co/datasets/OpenPipe/h
9.
by neutralino1 2y ago
Oh you cannot actually view the raw embedding vector, only the corresponding row. Any particular use-case to view the raw embedding?
10.
7 Different Prompting Techniques with Examples(axflow.dev)
1 points by neutralino1 3y ago|0 comments
11.
by neutralino1 3y ago
There are cures for hiccups!?
12.
by neutralino1 3y ago
Satya is buying popcorn.
13.
by neutralino1 3y ago
Price is the advantage.
14.
by neutralino1 3y ago
I often draw the parallel with cigarettes and alcohol. Kids need to produce an ID to purchase them. Sure they can fake it, but then they are breaking the law, and that still raises the barrier. But that's likely not enough. In addition
15.
by neutralino1 3y ago
Ideally yes, but the liabilities are likely too great at this time. So better start with the "hopeless", where no harm can be done, then go up from there.
16.
by neutralino1 3y ago
Congrats on the launch!
17.
by neutralino1 3y ago
I wouldn't go as far as to say business executives have good social skills. They are often ruthless, cunning, and deceiving, which makes them successful. If by good social skills you mean the ability to convince (read deceive) a lot of
18.
by neutralino1 3y ago
A lot of ads on this page.
19.
Comparing Software Developers with ChatGPT: An Empirical Investigation(arxiv.org)
2 points by neutralino1 3y ago|0 comments
20.
Bot or Human? Detecting ChatGPT Imposters with a Single Question(arxiv.org)
2 points by neutralino1 3y ago|0 comments
21.
Sematic (YC S22) is hiring their first enterprise sales representative(ycombinator.com)
1 points by neutralino1 3y ago
22.
by neutralino1 3y ago
Sematic | Enterprise Sales Representative | Full-time | Hybrid with office in SF | $100k-$150k base, equity, benefits Sematic (YC S22) is the open-source Continuous Machine Learning platform built by industry veterans (Cruise, Google, Micro
23.
Tesla Model 3 Hacked in Less Than 2 Minutes at Pwn2Own Contest(darkreading.com)
2 points by neutralino1 3y ago|0 comments
24.
by neutralino1 4y ago
Thank you – You are right that these are very important topics, and we also had to expend a lot of work at Cruise to scale training beyond single node. We had training jobs running over dozens of GPU nodes for many days. For example, we had
25.
by neutralino1 4y ago
> - How does it handle failure of individual tasks in the pipeline? At this time there are no handling of failures (Sematic is 6 weeks old :). In the near future we will have fault tolerance mechanisms: retries, try/except. > - W
26.
by neutralino1 4y ago
You are free to launch jobs on Spark clusters from any of your pipeline steps, whether they run locally or remotely. Be mindful that if you do so, the function that launched the job needs to also wait for its completion. In the future we wi
27.
by neutralino1 4y ago
You can certainly deploy the web app on one server, and run your pipelines on another (or the same). In this case, "locally" would mean that the pipeline and all its steps run on the same host machine. This is totally sufficient i
28.
by neutralino1 4y ago
As they say, naming things is one of the two hardest problems in Computer Science :) Your project is cool and has a cool name!
29.
by neutralino1 4y ago
Thank you! What we have seen done in the past is use things like pybind11 to expose C++ APIs in Python, which I guess is a similar concept to SWIG. If you are using build tools such as Bazel, you can even get the C++ compiled at run-time wh
30.
by neutralino1 4y ago
Thanks! I was a pretty heavy user of Rails in past jobs. They have a nice mix of good abstractions (Model/View/Controller), tooling (CLI, local web server), and best practices (how to name tables, fields, lifecycle timestamps, etc

More ›