druskacik

6 ms·

61.
by druskacik 1y ago
It's usually better to pay in the local currency than Euro. The currency still has to be exchanged somewhere, and the banks usually have better rates than the terminal operators.
62.
Why Doesn't OpenAI Build GPT Search Console?(robertdruska.com)
1 points by druskacik 1y ago|0 comments
63.
Ask HN: Bing Search API will be retired soon. What alternatives do you use?
2 points by druskacik 1y ago|1 comments
64.
by druskacik 1y ago
If someone wants to begin with Baudelaire, I would recommend his collection Le Spleen de Paris [0] - it's not poetry in its usual sense, but a collection of prose poems . I still remember picking the book randomly at our city library
65.
Direct chat with LLM from address bar in Firefox(robertdruska.com)
1 points by druskacik 1y ago|0 comments
66.
by druskacik 1y ago
Yes, the tags are not always completely relevant - they are topics for the range of posts the account posted. There's also a `main_topic`, which should be more relevant, but I did not add an option to filter by it yet.
67.
Show HN: Database of 1M Instagram influencers grouped into categories(infldb.com)
2 points by druskacik 1y ago|2 comments
68.
by druskacik 2y ago
That's what I was thinking. I hoped to see a price drop, but this does not change anything for my use cases. I was using gpt-4o-mini with batch API, which I recently replaced with mistral-small-latest batch API, which costs $0.10/
69.
EuroBERT: A High-Performance Multilingual Encoder Model(huggingface.co)
3 points by druskacik 2y ago|1 comments
70.
Claude as a Video Editor(robertdruska.com)
1 points by druskacik 2y ago|1 comments
71.
by druskacik 2y ago
Nice. One small bug: on small laptop screens, the lower parts of chessboard overflow out of view and it's impossible to scroll down to see them :)
72.
by druskacik 2y ago
Well, it's $2.19 per million output tokens even directly on deepseek platform. https://api-docs.deepseek.com/quick_start/pricing/
73.
by druskacik 2y ago
It's still crazy to me that 150K can be considered an unsatisfactory salary. I thought that must be enough anywhere in the world.
74.
by druskacik 2y ago
You underestimate how much being a soccer fan means, especially in countries like UK.
75.
by druskacik 2y ago
How many Mistral puns are there? The benchmarks look promising, great job, Mistral.
76.
by druskacik 2y ago
Yes! This library is great and definitely helps, but I still had problems with performance. For example, smaller models would still hallucinate when extracting a JSON field when the given field wasn't present in the text (I'd expe
77.
by druskacik 2y ago
Data extraction is definitely one of the most useful functions of LLM, however, in my experience a large model is necessary for a reliable extraction - I tested smaller, open-weights models and the performance was not sufficient. I wonder,
78.
by druskacik 2y ago
Thanks for the comment. Yes, I thought about something like that - I want to create a subpage for every playlist where users will be able to upvote or downvote specific tracks. I also plan to add option to upvote the whole playlist, to get
79.
Show HN: Book to Classical Music Soundtrack(illuminovel.com)
4 points by druskacik 2y ago|2 comments
80.
by druskacik 2y ago
Yes, aistudio.google.com is the way to go! You can upload some long documents such as books and try the long context. And the AI Studio is actually free, not many people know about this. As of the recall performance, I can't really spe
81.
by druskacik 2y ago
Mainly for its long context abilities. The other major LLMs peak at 128-256K context windows, Gemini models promise 1 million (2 million in case of 1.5 Pro).
82.
by druskacik 2y ago
But if you use it on the web, it becomes "open-source" - not by license, but for all practical purposes. Or am I wrong?
83.
by druskacik 2y ago
There is a method that could help immensely when answering questions like these. E.g. some of these question may be answered quite quickly using WikiData [0] (answer to question about the recipients of Medal of Freedom, query written with t
84.
by druskacik 2y ago
Okay, this already exists, nice to know: https://www.wikishark.com/?text_search=&values=4628128%2C457...
85.
by druskacik 2y ago
Hm, very interesting. If this would be packaged into a neat web UI, I can imagine it becoming an open alternative to Google Trends.
86.
Show HN: Generate book character illustrations with AI (Gemini + Flux)(illuminovel.com)
1 points by druskacik 2y ago|0 comments
87.
by druskacik 2y ago
There's a project [0] that parses Commoncrawl data for various schemas, it contains some interesting datasets. [0] http://webdatacommons.org/
88.
by druskacik 2y ago
How were Firefox users marginalized by X?
89.
Web Data Commons(webdatacommons.org)
1 points by druskacik 2y ago|0 comments
90.
by druskacik 3y ago
Being shredded is definitely not a rule among the top football players. Case in point, Harry Kane: https://www.reddit.com/r/coys/comments/9orbvv/kanes_dad_bod_...

More ›