woodson

7 ms·

91.
by woodson 2y ago
What if you bought it in another country? (Directly from Google’s pixel store in that country.) Out of luck, that’s what.
92.
by woodson 2y ago
Shame that there’s often a delay of 1-2 months between the trades and when they’re disclosed, so the first option likely isn’t going to work.
93.
by woodson 2y ago
That’s the world we live in today. Under many countries’ privacy laws, it’s not legal to sell PII to a third party that you collected for a specific other purpose (e.g., fulfilling the primary purpose of the app). The problem is that they d
94.
by woodson 2y ago
I think it's just "-x". It should select the best audio quality automatically.
95.
by woodson 2y ago
Thanks for the added context on the builds! As "foreign" BW player and fellow speech processing researcher, I agree shallow contextual biasing should help. While not difficult to implement, most generally available ASR solutions d
96.
by woodson 2y ago
I also liked his interview project: https://youtube.com/@davidlynchpresentsIP
97.
by woodson 2y ago
Python packages (distributed via the Python package index, PyPI) can also be general binary packages. Try pip install cmake, for example.
98.
by woodson 2y ago
If you have labelled data, you can try linear discriminant analysis (LDA; also known as canonical discriminant analysis), which maximizes the between-class variance while minimizing the within-class variance to best separate different class
99.
by woodson 2y ago
Portland is the same. There used to be Coffee Time on NW 21st, but that now closes at 6pm, too.
100.
by woodson 2y ago
Query keyword expansion works quite well for that without semantic search (although it can reduce precision).
101.
by woodson 2y ago
Sounds great! Are you (or anyone reading this) interested in doing a Stammtisch at Stammtisch?
102.
by woodson 2y ago
> Roles tied to research institutions cap-exempt > - Organizations where fundamental research is a key activity now qualify > - Startups can hire researchers (AI, health, hardware) year-round That’s a good change. I’ve seen ML re
103.
by woodson 2y ago
To be fair, one would use more than 6 steps with the original Hunyuan model, so perhaps that’s why they’re so blurry. But that’s even slower.
104.
by woodson 2y ago
I assume any company offering VoIP services that interact with phone numbers (Direct inward/outward dialing, DID etc) is potentially included. E.g., virtual PBX, Twilio and so on.
105.
by woodson 2y ago
But usually only upon acceptance for publication (i.e., an editor of a scientific journal has decided to accept the manuscript, after peer review and usually multiple revisions based on said review). Preprint papers might as well be some cr
106.
by woodson 2y ago
There are lots and lots of models, covering various use cases (e.g., on device, streaming/low-latency, specific languages). People somehow think OpenAI invented audio transcription with whisper in 2022 when other models exist and have
107.
by woodson 2y ago
That’s just using WebRTC-VAD, right?
108.
by woodson 2y ago
All these papers doing "research" on how to better prompt ChatGPT would be unpublishable then, given that API access to older models gets retired, so the findings of these papers can no longer be reproduced. (I agree with you in p
109.
by woodson 2y ago
As a CA, how does one accidentally issue a certificate for google.com? I mean, is there a scenario that isn't malicious?
110.
by woodson 2y ago
The strange thing is that the same thing is happening in the US. A story of someone’s Google, PayPal, Instagram etc account getting banned for no reason gets posted every day. Customer service means posting on social media and hoping it get
111.
by woodson 2y ago
Using a built-in vector extension is convenient if you want to integrate vector similarity (“semantic search”) with faceted search. Some vector stores (e.g., qdrant) support vector attributes that can be matched against, though. As mentione
112.
by woodson 2y ago
They got mad because you got Recall in an update, no matter whether you wanted it or not, and after another update you couldn’t uninstall it anymore. No choice.
113.
by woodson 2y ago
I say “I work in machine learning, what kids these days call AI” ;).
114.
by woodson 2y ago
It depends. If you're using Python with numpy>=2.0.0 (and macOS>=14) then you should benefit greatly from Apple's Accelerate implementation of BLAS/LAPACK routines which are behind most linear algebra operations. I&#x27
115.
by woodson 2y ago
There’s also John Markoff’s history of the relationship of human and intelligent machines, artificial intelligence vs intelligence augmentation, etc. ( https://www.goodreads.com/book/show/23460922-machines-of-lov...
116.
by woodson 2y ago
Unfortunately, it often hallucinates wrong parameters (or gets their order wrong) if there are multiple different APIs for similar packages. For example, there are plenty ML model inference packages, and the code suggestions for NVIDIA Trit
117.
by woodson 2y ago
Yes, its performance is rather poor and there can be a lot of headaches with caching (especially if you're using a file system that doesn't support reflinks). For large sharded datasets (e.g. WebDataset), you're better off wi
118.
by woodson 2y ago
There’s just not a single one-size-fits-all model/pipeline. You choose the right one for the job, depending on whether you need streaming (i.e., low latency; words output right when they’re spoken), run on device (e.g. phone) or server
119.
by woodson 2y ago
Their specifications include the technology used in telephony, such as encodings/codecs, link adaptation schemes, etc. They also publish reference implementations of many of the algorithms used.
120.
by woodson 2y ago
There are some approaches that use an LLM to generate “scripts” (you can think of them as a DSL) for composing/arranging media, essentially driving other models to generate parts of the media. One example is WavJourney: https:/&#

More ›