rfoo

8 ms·

121.
by rfoo 1y ago
What if I prefer to have a clone of me doing my coding, and then I throw my clone under the bus and start to (angrily) hyperfocus explore and change every piece to be beautiful? Does this mean I love coding or I hate coding? It's defin
122.
by rfoo 1y ago
Meanwhile, in a certain modern OS, unloading a library is too broken to the point that people are discouraged to do so... Try to unload GLib [0] from your process :p [0] https://docs.gtk.org/glib/
123.
by rfoo 1y ago
`Any` is the correct call. It could be: def f(i=0) -> None: if i is None: do_something() else: do_something_else() Yeah, I know it's retarded. I don't expect high quality code in a code base missing t
124.
by rfoo 1y ago
> so much from your macbook At least on cloud I can actually have hundreds of GiBs of RAM. If I want this on my Macbook it's even more expensive than my cloud bill.
125.
by rfoo 1y ago
There are a lot of examples about SQL in comments. In the SQL case you want something like: def process_template(template: Template) -> tuple[str, tuple]: sql_parts = [] args = [] for item in template: if isinstance
126.
by rfoo 1y ago
Google has been doing this since forever for recaptcha. And, to be fair, it seems to be fairly effectively for bot detection. https://github.com/neuroradiology/InsideReCaptcha > bots seem to be going a very differen
127.
by rfoo 1y ago
> I've already seen this story play out with low-code. The pitch was the same That's the point. Unlike low-code or no-code bullshit, this somehow worked. People "magically" want to design solutions, and are able to, n
128.
by rfoo 1y ago
And I pay Cursor only for autocomplete - this explains the difference I guess. I do sometimes use Composer (or Agent in recent versions), but it's being increasingly less useful in my case. Not sure why :(
129.
by rfoo 1y ago
> Cursor isn't providing their own models For use cases demanding the most intelligent model, yes they aren't. However, there are cases that you just can't use best models due to latency. For example next edit prediction,
130.
by rfoo 1y ago
Codeium's completion model is better than whatever GitHub Copilot has. For me it's Cursor > Codeium >>> Copilot. Yes, Copilot is that bad. And yes Codeium/Windsurf focuses on enterprise customers more. As GP said
131.
by rfoo 1y ago
Plus that they finally have someone else to blame, and it seems like people want to pretend it's all someone else's fault too. Probably don't even need to use their authoritarian cards.
132.
by rfoo 1y ago
In the same way how you manage a group of brilliant interns.
133.
by rfoo 1y ago
> I think 15M random read IOPS is actually quite fast 680x NVMe SSDs over 68 storage servers (so 68 CPUs) for just 15M (or 25M, tuned) random read IOPS is pretty underwhelming. The use cases where 3FS (or some other custom designs) shine
134.
by rfoo 1y ago
I'd estimate that there would be two orders of magnitude of difference in 4K random IOPS. If not three.
135.
by rfoo 1y ago
IMO they serve similar at a glance, but actually very different use cases. SeaweedFS is more about amazing small object read performance because you effectively have no metadata to query to read an object. You just distribute volume id, fil
136.
by rfoo 1y ago
3FS isn't particularly fast in mdbench, though. Maybe our FDB tuning skill is what to blame, or FUSE, I don't know, but it doesn't really matter. The truly amazing part for me is combining NVMe SSD + RDMA + supports reading a
137.
by rfoo 1y ago
> If you're in the European Economic Area, Switzerland, or the United Kingdom, the terms under "How Google uses Your Data" in "Paid Services" apply to all Services, including Google AI Studio and unpaid quota in
138.
by rfoo 1y ago
Before this change, 1000 512KB GET requests cost $0.0002 ($2e-4). After this change, 1000 512KB GET requests cost $0.00003 + $0.0006 / 1024*3 * 512 * 1024 * 1000 = $0.00003 ($3e-5) + $0.00029296875 (2.93e-4, dominates the cost) = $0.00
139.
by rfoo 1y ago
> Codebase Divergence: Our engine is based on an early fork of vLLM from over a year ago If you are in the same boat you'll see how much changed in vLLM compared to one year ago. Also, this meant that they haven't rebased for o
140.
by rfoo 1y ago
If all your GET requests are 512KB in response, this "price reduction" is effectively a price raising. GET requests go up to $0.000323 per 1K from $0.0002 per 1K.
141.
by rfoo 1y ago
tl;dr "we had our vLLM fork and it's unmaintainable now; guess we are going to rebuild it, in the public this time"
142.
by rfoo 1y ago
> a "no malicious actors" assumption in its design in order to make the performance/scaling guarantees it does Didn't think deep into it, could this be solved with billing designs with more nuance?
143.
by rfoo 1y ago
It is well possible by changing 50 lines of code in a 10GB project you have to rebuild the entire project, if everything (indirectly) depends on what you just changed.
144.
by rfoo 1y ago
I think the joke here is that somehow management refused to sell Colossus (which is such an obvious nice product just like BigQuery) before and it takes "AI" to convince them.
145.
by rfoo 1y ago
I don't. I have no problem not running open-weight models myself because there's an efficiency gap of two orders of magnitude between "pretend-I-can" solution and running them on hundreds of H100s for high thousands of u
146.
by rfoo 2y ago
> see o3-mini's release date & pricing vs R1's This backfires though, if OAI released o3-mini before DeepSeek-R1, R1 would be a lot less impactful.
147.
by rfoo 2y ago
That's about the same number for DeepSeek-V3. If you count in fp8 MFU is about 20%. MoEs are hard. That could also be why they did fp8. If we use theoretical performance of bf16 as baseline (I know this makes few sense, but for compare
148.
by rfoo 2y ago
From model cards, suggested system prompt: > You are Llama 4. Your knowledge cutoff date is August 2024. You speak Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai, and Vietnamese. Respond i
149.
by rfoo 2y ago
The point is, it isn't marginally better. I agree the setup is not a demonstration of intelligence, but the difference is pretty significant. Not to mention that on conventional benchmarks Llama 405B is usually worse than GPT-4o.
150.
by rfoo 2y ago
I think the most interesting result [0] is, compared to our current benchmarks, on which scaling law is showing diminishing returns, what they did managed to tell apart large language models (Llama 405B, GPT-4.5) from not-so-large LMs. This

More ›