leumon

7 ms·

91.
by leumon 1y ago
Just like for llms for the base model you need quantity not quality. It just needs to learn how to correctly predict the next frame.
92.
by leumon 1y ago
“Thirty seconds after take off, there was a loud noise and then the plane crashed. It all happened so quickly,” said Ramesh, speaking to the Hindustan Times. He said he “impact injuries”, including bruising on his chest, eyes and feet but
93.
by leumon 1y ago
Possibly to get more training data.
94.
by leumon 1y ago
Except that's not true: an auto shutdown only occurs at 47.5 Hz, at least in Germany, some countries have the limit even lower: https://de.m.wikipedia.org/wiki/Unterfrequenz
95.
by leumon 1y ago
> Even though these ad algorithms are not nearly perfect (try to pay attention to how often you are served ads that are entirely irrelevant to your interests), the simple fact that they are so eerily correct even some of the time is the
96.
by leumon 2y ago
elo ranking based on user preferences. you can vote on prompts and submit new ones in the 'video arena' tab
97.
by leumon 2y ago
In my opinion qwq is the strongest model that fits on a single gpu (Rtx 3090 for example, in Q4_K_M quantization which is the standard in Ollama)
98.
by leumon 2y ago
well already in 2013 ocr systems used in xerox scanners (turned on by default!) randomly altered numbers, so its not an issue only occuring in llms.
99.
by leumon 2y ago
He should've tested 4.5. This model is hallucinating much less than any other model.
100.
by leumon 2y ago
this model does have a niche use-case: since its so large it does have a lot more knowledge and hallucinates much less. for example as a test question I asked it to list the best restaurants in my small town. and all of them existed. none o
101.
by leumon 2y ago
Unfortunately the app is not compatible with Android 15.
102.
by leumon 2y ago
I've found gemini-1206 to be best. and we can use it free (for now), in google's aistudio. It's number 1 on lmarena.ai for coding, and generally, and number 1 on bigcodebench.
103.
by leumon 2y ago
While it now can read an analog clock from an image (even with seconds), on some images it still doesn't work https://i.imgur.com/M2JouZs.png
104.
by leumon 2y ago
Using comfyui with the official flux workflow is easy and works nicely. comfy can also be used via API.
105.
by leumon 2y ago
I think it depends on the amount of the radiation. While non-ionizating radiation doesn't have the ability to remove electrons, it still can heat up surfaces. Microwave ovens for example use non-ionizing radiation to heat up food, but
106.
by leumon 2y ago
Maybe musk should bring back the onion link to Twitter that he purged.
107.
by leumon 2y ago
I've noticed much higher level of service in the US compared to Europe were tips are not expected as much Might be true, but I don't think that's necessarily a good thing. If someone is forced to be excessively friendly&#x2
108.
by leumon 2y ago
Just give us airtags in icloud.com/find please, so that I can use them while having an android phone. I don't need apple maps in browser.
109.
by leumon 2y ago
It's difficult to do this for a video (and probably wouldn't look that nice)
110.
by leumon 2y ago
They are only taken over in certain edge cases, and even then the remote operator is only placing the next destination, he is not actually remotely "driving" the vehicle. More on this here: https://waymo.com/blog&#
111.
by leumon 2y ago
I wonder what would happen if the customer is very happy and having an enjoyable conversation. Would the AI remove the emotions here too, resulting in a completely monotonic voice?
112.
by leumon 2y ago
Glad to hear that apple did not get away with this. I just hope that they can end up actually allowing sideloading, meaning we can install any ipa, just like you can run any executable on a mac. Is the mac less secure because of this? No.
113.
by leumon 2y ago
I think it's not actually in the training data. The llm is just using RAG, meaning it gets the top search results for that query and based on that generates a text. Kinda like perplexity but apparently worse then it.
114.
by leumon 2y ago
I think it is more then a simple tts engine. At least from the demo, they showed: It can control the speed and it can sing when requested. Maybe its still a seperate speech engine, but more closely connected to the llm.
115.
by leumon 2y ago
You can request to dissociate them from your account. (better then nothing)
116.
by leumon 3y ago
Mistral did it too when they released their first open model. They just posted a magnet link on Twitter.
117.
by leumon 3y ago
For logical stuff, GPT-4 is still superior, especially for more complex stuff. I like Claude for creating more simple stuff that doesn't require that much reasoning because it generally writes very detailed output.
118.
Show HN: LectureRecorder – Switch Slides While in Lecture to Record Audio+Slides(github.com)
2 points by leumon 3y ago|0 comments
119.
by leumon 3y ago
Things like ShotSpotter are only fighting the symptoms. As an european I find it ridiculus to hear that such thing even exists. Why would we not simply ban the guns themselves?
120.
by leumon 3y ago
Funny that they say "Its models are “open source”, meaning technical details will be released publicly.". In fact, they aren't only releasing the technical details, they are also releasing the model weights themselves.

More ›