operator-name

8 ms·

61.
by operator-name 3y ago
The capitalisation has just made me realise that it's "Photo Pea" and not "Photo-pe-a".
62.
by operator-name 3y ago
There's also been a greater focus on reproducibility: https://guix.gnu.org/en/blog/2023/the-full-source-bootstrap-...
63.
by operator-name 3y ago
For some reason it's actually bundling both LLaMA 13b (24.5GB) and Ministral 7b (13.6GB), but only installed Ministral 7b. I have a 3070ti 8GB, so maybe it installs the other one if you have more VRAM?
64.
by operator-name 3y ago
Having installed this, this is an incredibly then wrapper around the following github repos: https://github.com/NVIDIA/trt-llm-rag-windows https://github.com/NVIDIA/TensorRT-LLM It's quite a
65.
by operator-name 3y ago
The source is available, minus the installer. You could always use the base repo after verifying it: https://github.com/NVIDIA/trt-llm-rag-windows
66.
by operator-name 3y ago
Yeah, seems a bit odd because the TensorRT-LLM repo lists Turing as supported architecture. https://github.com/NVIDIA/TensorRT-LLM?tab=readme-ov-file#pr...
67.
by operator-name 3y ago
I found some official benchmarks for enterprise GPUs, but no comparison data. I couldn't find any benchmarks for commercial GPUs. https://nvidia.github.io/TensorRT-LLM/performance.html
68.
by operator-name 3y ago
Last time I used it, LM Studio doesn't include RAG.
69.
by operator-name 3y ago
This looks quite cool! It's basically a tech demo for TensorRT-LLM, a framework that amongst other things optimises inference time for LLMs on Nvidia cards. Their base repo supports quite a few models. Previously there was TensorRT for
70.
by operator-name 3y ago
This is a tech demo for TensorRT, which is ment to greatly improve inference time for compatible models.
71.
by operator-name 3y ago
Honestly, this is a really cool demonstration of how their foveated rendering works.
72.
by operator-name 3y ago
It seems to come with Google One (2TB), so with that factored in its actually quite competitive.
73.
by operator-name 3y ago
When you registered for 2FA it will have given you backup tokens to be written down. They're designed for this exact situation. Given the number of software projects that use github as their canonical distribution platform and the num
74.
by operator-name 3y ago
Note that cheerpj is not open source: https://labs.leaningtech.com/cheerpj3/licensing
75.
by operator-name 3y ago
Announcement from the Rust side: https://news.ycombinator.com/item?id=39263750
76.
by operator-name 3y ago
It's interesting that they use liquid metal for the CPU thermal interface. Unlike normal liquid metal it seems to be solid at room temperature, which I'm guessing was for ease of (re)application. > We’re using Coollaboratory’s
77.
by operator-name 3y ago
Gandi lists lol.meme as a premium domain for over $800. We're also in the early access period where even normal domains cost thousands of dollars to reserve.
78.
by operator-name 3y ago
https://jarv.is/notes/cloudflare-dns-archive-is-blocked/
79.
by operator-name 3y ago
There is an equally good solar eclipse in April 2024 that crosses the country the other way around. https://science.nasa.gov/eclipses/future-eclipses/eclipse-20...
80.
by operator-name 3y ago
This is complete speculation, but the luminance difference between the red sprites and greyscale shirt is greater than that of the other colors.
81.
by operator-name 3y ago
This is beautiful. The simple paragraph accompanied with links to "The Dictionary of Obscure Sorrows" so eloquently expresses what I've been mulling on.
82.
by operator-name 3y ago
That was the decoy behind the secondary infection vector, not the motivation.
83.
by operator-name 3y ago
> The shellcode used in this exploit is constructed in a similar manner to shellcode observed in previous North Korean exploits. At minimum the payload.
84.
by operator-name 3y ago
Imgur mirror: https://i.imgur.com/TgfKL9a.jpg
85.
by operator-name 3y ago
Same topic, but a news article: https://news.ycombinator.com/item?id=37311239
86.
by operator-name 3y ago
https://knowyourmeme.com/memes/cheems
87.
Tor Release 0.4.8.1-Alpha Includes Onion Service Proof-of-Work(forum.torproject.org)
4 points by operator-name 3y ago|1 comments
88.
by operator-name 3y ago
0.4.8.3-rc released 5 days ago, and 0.4.8 stable is set to release in the coming week or so. https://forum.torproject.org/t/release-candidate-0-4-8-3-rc/...
89.
by operator-name 3y ago
I've not looked at the specific implementation here, but Tor's implementation[0] includes a dynamic difficulty scaling. When under attack, legitimate users will experience a moderate delay whilst attackers will need to scale their
90.
by operator-name 3y ago
I think I've reached an understanding, but do correct me if I'm wrong. lim players -> inf mean wealth -> starting_wealth * (win_chance * (1 + win_gain) + lose_chance * (1 - lose_loss)^rounds) So in their example

More ›