Worthy thoughts.
Worthy thoughts. But you should think of the Icelandic eth ("ð") as a variation on a "d", not an "o". :-) Not that I know any Icelandic, but a capital eth looks like this: "Ð". More info: https://en.wikipedia.org/wiki/Eth EDIT: As for "クッキー": dealing with this is mostly straightforward, if you're willing to allow multiple ASCII characters for a single unicode character. Most of the katakana (which these are), along with the hiragana, have a standard romanization. For the first three here, the romanizations are (if I'm not mistaken) "ku", "tu", and "ki". The dash-looking thing lengthens the last vowel, so this is "kutukii". That probably requires more intelligence than you're wanting to bake into this thing, but I think "kutuki-" wouldn't be bad. Korean characters similarly have standard romanizations (e.g., "원" = "won"). Figuring out all of the above would take some work, but it could easily be crowd-sourced, and it would only have to be done once. The zillions of Chinese characters are the problem. These could have different romanizations depending on which of the Chinese languages they're being used for, and often multiple possible romanizations when used for Japanese (in which case they are "kanji"). So there might be no good solution for these.
Posted by ggchappell on Aug 29, 2014, 11:38 PM
Tags
Related issues
- Not sure what you suggest to do.
- Why is this not yet published on the official website and why is there no CVE?
- The biggest problem with having an LLM do this, from what I've observed of at least frontier models, is that the way they are trained...
- The biggest problem with MOND is that it was hand-crafted with additional parameters simply to fit the data.
- (My comment is about detached houses).
- The biggest problem with this is that I can't run my company's own internal benchmarks against this API, because they will be trained int...
Raise real-world problems, find people facing the same issue, and discuss solutions.9,000 issues0 comments
hackernews (Other)
Not sure what you suggest to do.
NewWhy is this not yet published on the official website and why is there no CVE?
NewThe biggest problem with having an LLM do this, from what I've observed of at least frontier models, is that the way they are trained...
The biggest problem with MOND is that it was hand-crafted with additional parameters simply to fit the data.
(My comment is about detached houses).
The biggest problem with this is that I can't run my company's own internal benchmarks against this API, because they will be trained int...
> steer Claude away from over-engineering parts of the system in response to hypothetical concerns.
Why is there no comparison with FLUX 3 Image?
Worthy thoughts.
Worthy thoughts. But you should think of the Icelandic eth ("ð") as a variation on a "d", not an "o". :-) Not that I know any Icelandic, but a capital eth looks like this: "Ð". More info: https://en.wikipedia.org/wiki/Eth EDIT: As for "クッキー": dealing with this is mostly straightforward, if you're willing to allow multiple ASCII characters for a single unicode character. Most of the katakana (which these are), along with the hiragana, have a standard romanization. For the first three here, the romanizations are (if I'm not mistaken) "ku", "tu", and "ki". The dash-looking thing lengthens the last vowel, so this is "kutukii". That probably requires more intelligence than you're wanting to bake into this thing, but I think "kutuki-" wouldn't be bad. Korean characters similarly have standard romanizations (e.g., "원" = "won"). Figuring out all of the above would take some work, but it could easily be crowd-sourced, and it would only have to be done once. The zillions of Chinese characters are the problem. These could have different romanizations depending on which of the Chinese languages they're being used for, and often multiple possible romanizations when used for Japanese (in which case they are "kanji"). So there might be no good solution for these.
Comments (0)
No comments yet. Start the discussion.