resiros

9 ms·

61.
by resiros 2y ago
yes, please.
62.
by resiros 2y ago
Congrats on the launch! The idea sounds very interesting on paper. The tricky part though is the reward function. Providing finetuning as a service works because the friction with finetuning is operational (getting the GPUs, preparing the t
63.
by resiros 2y ago
Ditto, living there too. South Berlin is unfortunately desolate, with more graveyards than coffee shops. It would be awesome if it would change.
64.
by resiros 2y ago
I am a founder who lives 10 minute walk from there. I never heard about it before. It's exciting and all, but I am quite certain no member of our team would be happy to commute to Mariendorf for work :D
65.
What We Learned Building a Prompt Management System(agenta.ai)
2 points by resiros 2y ago|0 comments
66.
by resiros 2y ago
Arabic speaker here. No, it's perfect.
67.
by resiros 2y ago
Check out agenta.ai it's OSS and provide tools for prompt management, evaluation and observability (with otel). I think it's easy to get started.
68.
by resiros 2y ago
The discord invite link ( https://discord.gg/xJJMRaWCRt ) in ( https://www.latent.space/p/community ) is invalid
69.
by resiros 2y ago
One has to respect his procrastination skills. He would go to learn quantum mechanics just not to complete the game of thrones books.
70.
by resiros 2y ago
That's a way of looking at it. Another way is looking at it inwardly, as a journey of self. Personally, a PhD for me was being punched in the face every morning (i.e., experiment failed) for years, yet getting up to try again. It'
71.
by resiros 2y ago
I would argue that it's right call: 1) it's in the world's best interest. I am running llama locally on laptop, and the ability to have the distilled world's knowledge at your fingertips will generate much much more valu
72.
by resiros 2y ago
The author discusses the problem from the point of engineering, not from business. When you look at it from business perspective, there is a big advantage of not waiting, and using whatever exists right now to solve the business problem, so
73.
by resiros 2y ago
I went ahead and bought the book from interest in helm (quite the beginner in k8 and all). Unfortunately it's not very far from an AI generated text about helm. It's 28 pages of text (not far from a long blog post), it does not ha
74.
by resiros 2y ago
One missing in the list below is Agenta ( https://github.com/agenta-ai/agenta ). We're oss, otel compliant with stronger focus on evals and the enabling collaboration between subject matter experts and devs.
75.
by resiros 2y ago
I don't think it's worse, just that the standards moved up, the expectations are now much higher. Sam forgot YC's advice to launch fast.
76.
by resiros 2y ago
Could be, but he does not strike as someone who is looking for fame. Plus the whole discussion about why he would like to move to SF but can't seems pretty authentic.
77.
by resiros 2y ago
>Wait if you’re doing $900-1000/month and you’re sustaining yourself on that, that must mean you’re sustaining yourself on less than $12,000 a year. What is your lifestyle like at $12K?" >I live in the middle of nowhere. I
78.
by resiros 2y ago
Check out Agenta ( https://github.com/agenta-ai/agenta or agenta.ai), it allows you to debug as in you have a playground where you can see the traces for each call you make and run evals on your workflow
79.
by resiros 2y ago
I think the innovation is using caching as so to make the cost of the approach manageable. The way they implemented it is that each time you create a chunk, you ask the llm to create an atomic chunk from the whole context. You need to do th
80.
by resiros 2y ago
He did not. He stopped his jet to refuel in France, and they issued it in the mean time
81.
by resiros 2y ago
I think the issue is that many of these metrics (e.g. RAGAS) are LLM as a judge metrics. These are very far from reliable. Making them reliable is still a research problem. I've seen a couple of startups training their own LLM judge mo
82.
by resiros 2y ago
Right now we have our own instrumentation but we're working towards Otel compatibility.
83.
by resiros 2y ago
Not OP here (but building in the same space). The reason you instrument LLM data is usually to improve quality/speed of your applications. The tools to extract the insights to enable that, and the integration with your LLM experimenta
84.
by resiros 2y ago
Just wanted to say great work on standardizing otel for LLM applications ( https://github.com/open-telemetry/semantic-conventions/tree/... ] and opensourcing OpenLLMetry. We're also building in this space,
85.
Ask HN: Why is Meticulous (YC S21) still hiring to eliminate E2E tests?
4 points by resiros 2y ago|2 comments
86.
Is AI a Silver Bullet?(ian-cooper.writeas.com)
2 points by resiros 2y ago|0 comments
87.
by resiros 2y ago
Agenta | Hybrid (Berlin) | Full Time Agenta is an open-source LLM developer platform that provides all the necessary tools for rapidly building reliable LLM applications. Features include prompt management, evaluation, human feedback, and o
88.
by resiros 2y ago
Lots of assumption here. First, that we will only be training on text data, if we take into considerations all the videos and audios shared I am quite sure we would have one or two orders of magnitude more of data. Second, that it even matt
89.
by resiros 2y ago
If you look into the instructor code( https://github.com/jxnl/instructor/blob/06a49e7824729b8df1f7... ). Here is the core code snippet they use: message = dedent( f""&quo
90.
by resiros 2y ago
Thanks for sharing the link, but no agenta is not a library that can help with getting structured outputs from LLMs (at least not in the way discussed in the parent comment). It's a prompt management, evaluation and observability platf

More ›