tokencost
csswg-drafts
tokencost | csswg-drafts | |
---|---|---|
6 | 72 | |
1,170 | 4,344 | |
81.8% | 1.0% | |
9.1 | 9.9 | |
7 days ago | 5 days ago | |
Python | Bikeshed | |
MIT License | GNU General Public License v3.0 or later |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
tokencost
-
Why is Everyone into Indie Development? - FAV0 Weekly Issue 004
Library for Estimating Token Costs
-
Show HN: Token price calculator for 400+ LLMs
I really appreciate your engagement here and think it has great value on a personal level, but the length and claims tend to hide two very obvious, straightforward things:
1. They only support GPT3.5 and GPT4.0. Note here: [1], and that gpt-4o would get swallowed into gpt-4-0613.
2. This will lead to massive, significant, embarrassingly large error in calculations. Tokenizers are not mostly the same, within 10% error.
1. responsive to ex. "It's not just C100K though. It is for a few models [0]",
The link is to Tiktoken, OpenAI's tokenization library. There are literally more than GPT3.5 and GPT4.0 there, but they're just OpenAI's models, no one else's, none of the others in the long list in their documentation, and certainly not 400.
every single one of them is for a deprecated model, not served anymore, except c100k and o200k. As described above and shown in [1], their own code kneecaps the o200k and will use c100k"
2. Let me know what you'd want to see if you're curious the 30%+ error thing. I don't want to go to the trouble to guess at a test suite, then run one, that would make you confident you need to revise a prior that there's only +/- 10% difference between arbitrary tokenizers. I will almost assuredly choose one that isn't comprehensive enough, with your input.
For context, I run about 20 unit tests, for each of the big 5 providers, with the same prompts, to capture their input and output token counts to make sure I'm billing accurately.
Just to save you time, you won't be able to talk me down to "eh, good enough!" --- It *matters*, if it didn't, they'd be much more up front about the truth. Every single sign around the library is absolutely damning, and triangulates somewhere between lying and naivete. From the marketing claiming 400+, to the complete lack of note of these extreme* caveats in any documentation, the only thing being what I understand is a warning log.
[1] https://github.com/AgentOps-AI/tokencost/blob/e1d52dbaa3ada2...
- Show HN: Easy token counting and price calculation for LLMs
csswg-drafts
-
Inline if() in CSS 🤔: What You Need to Know About This Upcoming Feature
Lets see how this will work ( according to the proposal ):
-
Why is Everyone into Indie Development? - FAV0 Weekly Issue 004
Github Issues Discussion
-
Help us invent CSS Grid Level 3, a.k.a. "Masonry" layout – WebKit
For more background, and some detailed discussion of the opposite argument ("display: masonry" over "display:grid"+"grid-template-rows: masonry") see https://github.com/w3c/csswg-drafts/issues/9041
-
Chrome Dev: High Definition CSS Color Guide
The tracking issue: https://github.com/w3c/csswg-drafts/issues/8659
As noted there, okHSL/HSV keeps the perceptual uniformity by removing some peaks beyond the geometric limit of HSL/HSV, and it is unclear whether it is what users do expect or not.
-
Announcing Winduum 1.0 - Framework agnostic component library for TailwindCSS
The idea is that you should be able to set accent color via accent-color CSS property. It is discussed that there should be access to the color value of this property, e.g. via AccentColor or AccentColorText.
-
Learn CSS Layout the Pedantic Way
What do you mean by "official documentation"? The specification [1]? MDN [2]?
[1] https://drafts.csswg.org/
[2] https://developer.mozilla.org/en-US/docs/Web/CSS
The former is not meant as a learning resource for new web devs and the latter usually has information about the "baseline" support ond browser compatibility tables.
-
CSS WG resolved to officially work on native custom functions and mixins
The link corresponding to the actual submission title (“CSS WG resolved to officially work on native custom functions and mixins”):
https://github.com/w3c/csswg-drafts/issues/9350#issuecomment...
> RESOLVED: Start ED of css-mixins for CSS Custom Functions and Mixins
-
Weird things engineers believe about Web development
Recently I was reading the Learn CSS the pedantic way book and the definition for inline boxes did not match the way that anonymous block boxes were generated when an inline-level element had a block-level element as its child. So I went looking elsewhere for a more appropriate definition for that case and found this issue on standards: https://github.com/w3c/csswg-drafts/issues/1477 It was really interesting to know that I was not the only one confused. My question was: Does the inline-box generated by the inline-level element contains the box generated by the block-level child or there wasn't an inline-box that was a parent of them all but there were 2 siblings inline-level boxes of the block-level box that were wrapped in another anonymous block boxes? Reading that issue I got to know the concept of fragments, which I did not know browsers had. But the issue seems to suggest that the box tree for this case should have the inline-box as being a parent of the block-box. Which led me to another question, in that case, if I apply a border to the parent inline-level element, shouldn't it apply to the overall box that is generated (it does not)? The answer is that borders between block-boxes and inline-level boxes should not intersect but that is really difficult to derive from reading the standards alone. Anyway it was headache-inducing trying to learn the box-model pedantically :)
-
CSS Is Fun Again
With all the recent CSS improvements I still miss the possibility to have working transition to "height:auto". The issue [1] on csswg-drafts is the most upvoted one. At least we can now use css grid and track sizes transitions, but it's far from intuitive, transition for "height:auto" should just work.
[1]: https://github.com/w3c/csswg-drafts/issues/626
- Proposed "au" unit for CSS provides for styling on an astronomical scale
What are some alternatives?
openai-messages-token-helper - A utility library for dealing with token counting for messages sent to an LLM (currently OpenAI models only)
Modernizr - Modernizr is a JavaScript library that detects HTML5 and CSS3 features in the user’s browser.
llm_utils - Utilities for Llama.cpp, Openai, Anthropic, Mistral-rs.
open-props - CSS custom properties to help accelerate adaptive and consistent design.
litellm - Call all LLM APIs using the OpenAI format. Use Bedrock, Azure, OpenAI, Cohere, Anthropic, Ollama, Sagemaker, HuggingFace, Replicate (100+ LLMs)
WHATWG HTML Standard - HTML Standard
anthropic-tokenizer - Approximation of the Claude 3 tokenizer by inspecting generation stream
modern-font-stacks - System font stack CSS organized by typeface classification for every modern operating system
container-query-polyfill - A polyfill for CSS Container Queries
rellax - Lightweight, vanilla javascript parallax library
Rotativa - Rotativa, /rota'tiva/. Make Pdf from Asp.Net MVC. Available on Nuget https://www.nuget.org/packages/Rotativa
standards-positions