nikhilsimha

7 ms·

31.
by nikhilsimha 3y ago
We did and gave up due to scalability limitations. Fundamentally most of the computation needs to happen before the read request is sent.
32.
by nikhilsimha 3y ago
It opens for me in incognito mode - albeit with a large popup that I had to close.
33.
by nikhilsimha 3y ago
There were two main drivers - - inability to back-test new real-time features. People were forced to log-and-wait to create training sets for months. Chronon reduces this to hours or days. - the difficulty of creating the lambda system (bat
34.
by nikhilsimha 3y ago
I agree with you - about the sentiment around the GenAI megaphone. FWIW, Chronon does serve context within prompts to personalize LLM responses. It is also used to time-travel new prompts for evaluation.
35.
by nikhilsimha 3y ago
Feathr from linkedin is the closest. But there doesn't seem to be much recent activity on the project.
36.
by nikhilsimha 3y ago
By mutable data do you mean - change data coming from OLTP databases? If yes, we do this via the EntitySource api. https://www.chronon.ai/authoring_features/Source.html#stream...
37.
by nikhilsimha 3y ago
I am with you on this one.
38.
by nikhilsimha 3y ago
Offline is pretty easy to get started with. It should take less than a week to set it up for new use-cases across the company. (You can begin building training-sets if offline is setup) Online is a bit more involved - you need a month or mo
39.
by nikhilsimha 3y ago
Bighead is the model training and inference platform. Chronon is a full re-write of zipline with 1) a different underlying algorithm for time-travel to address scalability concerns. 2) a different serde and fetching strategy to address late
40.
by nikhilsimha 3y ago
Airflow is the current implementation since it is the paved path at airbnb. But we are open to accepting contributions for other orchestrators. Someone mentioned they wanted to add cadence support.
41.
by nikhilsimha 3y ago
we call this chaining. In the API itself - you could specify the chain links by specifying the source. To be precise - a GroupBy(aggregation primitive) can have a Join(enrichment primitive) as a source. To rephrase, you can enrich first and
42.
by nikhilsimha 3y ago
re: scalability benchmarks - we plan to publish more benchmark information against publicly available datasets in the near future. re: human-in-the-loop workflows - do you mean labeling?
43.
by nikhilsimha 3y ago
the ability generate training sets against historical inferences to back-test new features another one is the focus on pushing as much compute to the write-side as possible (within Chronon) - specially joins and aggregations. OLAP databases
44.
by nikhilsimha 3y ago
Author. Happy to answer any questions.
45.
by nikhilsimha 3y ago
I like this translation: a.co/d/gt8dwbT
46.
by nikhilsimha 3y ago
Avadhuta Gita - a treatise on the non-dual nature of reality and consciousness. The allegation is that - the boundary between the “I” and “else”, that the “I” so strongly attaches itself to, is an illusion. The other claim is that, with med
47.
by nikhilsimha 3y ago
Brutal
48.
by nikhilsimha 3y ago
Llama 2 models were trained on slurm
49.
by nikhilsimha 3y ago
Never heard a union be called a “vertical join” before.
50.
by nikhilsimha 3y ago
cannot recommend sicp enough: https://mitp-content-server.mit.edu/books/content/sectbyfn/b... uses scheme to illustrate how to think about computation abstractly.
51.
by nikhilsimha 3y ago
/s ?
52.
by nikhilsimha 3y ago
France?
53.
by nikhilsimha 3y ago
This is one of the best articles I have read in a long long time!
54.
by nikhilsimha 3y ago
/s in the end? (genuine question)
55.
by nikhilsimha 3y ago
Very well done!
56.
by nikhilsimha 3y ago
And they worry about LLMs hallucinating
57.
by nikhilsimha 3y ago
I hope they take comfort in the fact that this is open-source.
58.
by nikhilsimha 3y ago
You could vote for the candidates (senators, state and national, governors and presidents) that push for nuclear, building-heating-tech regulations, carbon taxes on food production and tell your friends and family about it. If you have the
59.
by nikhilsimha 3y ago
It does get easier with repetition (mentally and socially). I do a water fast on thursdays (~36 hrs). I find that to be much easier than intermittent fasting mentally somehow.
60.
by nikhilsimha 4y ago
What can one do to effectively support these laws?

More ›