gunalx

9 ms·

121.
by gunalx 9mo ago
Well yes, but it isn't too cheap for how 8ld it is.
122.
by gunalx 9mo ago
Its more of a deepseek advert. The model here was the most incorrect one. (everyone else had half the errors or less) It dosent matter to be fast if you are wrong.
123.
by gunalx 9mo ago
Yep. Before gemma3 we where struggling with multilinguality on smaller European languages, and it is still one of the batter ones in that regard (even large open or closed models struggle with this to some extent). Gemma3 also is still pret
124.
by gunalx 9mo ago
its not yet, but gaining traction.
125.
by gunalx 9mo ago
new? They have done this a long time.
126.
by gunalx 9mo ago
This is the Max series models with unreleased weights, so probably larger than the largest released one. Also when refering to models, use huggingface or modelscope (wherever it is published) ollama is a really poor source on model info. th
127.
by gunalx 9mo ago
Microsoft and google seems hellbent on destroying classical email, by doing this crap. Their interoperability is also pretty bad.
128.
by gunalx 9mo ago
The indie dev market is a flip flop, I have seen many great apps only available on android as well.
129.
by gunalx 9mo ago
Yep, this is something end tl end models need to solve to be ideal I think. I hve seen a split brain architecture with one speaking and one thinking brain. If the thinking one could have some text tokens as output and input, to be able to r
130.
by gunalx 9mo ago
Voice actors are slo cooked. Some of the demos arguably sounded way better than a lot of indie voice-acting.
131.
by gunalx 9mo ago
not really though. The zed monetization seems to push towards selling tokens for full fledged models with good ide integration as a service. (They have let you run a custom tabcomplete for a little while)
132.
by gunalx 9mo ago
if you mean on the new ai hat with npu and integrated 8gb memory, maybe.
133.
by gunalx 9mo ago
^piper-tts exists.
134.
by gunalx 9mo ago
Never gibe perms to begin with. Anything the chatbot has access to fuckup it eventually will. So the problem is inherently flawed, but. Use db permissions with read only, and possibly only a set of prepared statements. Give it a useraccount
135.
by gunalx 9mo ago
Not to be pedantic. But facism is politics.
136.
by gunalx 9mo ago
I have seen post being used instead of get, because of having encrypted parameters by default.
137.
by gunalx 9mo ago
or instdad of ethanol have brewed methanol.
138.
by gunalx 9mo ago
I think it was that most of us coke had sirup, while mexican and most of the world uses cane sugar.
139.
by gunalx 9mo ago
Productivity and Cloud have a revenue of about 30B each while personal computing only was 13.5B (that includes windows Xbox and search + advertising) according to ms earnings report q4 25
140.
by gunalx 9mo ago
Its for times like this briar or similar applications are super useful. Also memory sticks. To bad distributing is not to simple.
141.
by gunalx 9mo ago
You can in fact apt install piper.
142.
by gunalx 10mo ago
Well, my ISP dosent support ipv6, and i get a non shared public ipv4, so no ipv6 here.
143.
by gunalx 10mo ago
Deepseek the model is not that censored. Deepseek the service is. So preaumably like openai and others, there is an additional model and filtering detecting misues or sensitive topics, and filtering the output.
144.
by gunalx 10mo ago
LLms are both faster, smarter and way dumber than a junior at the same time. They work faster, but more often make wrong assumptions without asking. Llms dont ask the stupid questions a junior might, but those questions are essential to get
145.
by gunalx 10mo ago
Not a us user, but ended up with us and uk layout, just because they where easier to find. (also works fine for programming)
146.
by gunalx 10mo ago
Also the free tier is sufficient for basically anything non power-user or enterprice.
147.
by gunalx 10mo ago
They say they dont train on your inputs. But their tos dosent prevent them from doing that. My guess is they do train on slightly altered/obfuscated user data.
148.
by gunalx 10mo ago
Until we end up spending trillions recreating the fab capacity of tsmc, they dont have a full monppoly (yet)
149.
by gunalx 10mo ago
Once had a complex model that would fully crash and lockup fusion, but once redone in openscad rendered after a little while. (badly designed + slow pc at tye time)
150.
by gunalx 10mo ago
There are tumbling services, where you for a fee can mix upp your transaction with lots of other users transactions to make it less obvious you where the one that transfered the credit to your burner wallet. Kepp in mind, tumblers have also

More ›