I'm a Data Scientist. For some time, I've been working on a library for feature engineering.
• GitHub: https://github.com/feature-express/feature-express
• Website: https://feature.express
It isn't yet complete, and I wouldn't consider it ready for production use or handling larger datasets. Here are some of its characteristics:
• Event-based workflows: Initially, everything is converted to an event format, ingested into an event store, and processed from there.
• In-memory: Both the event store and evaluation have been built from scratch.
• Written in Rust, but there's a Python package available.
• A DSL (Domain Specific Language) for defining aggregations, similar to SQL.
Why am I developing this? I've always found it challenging to build models based on time. These models can be surprisingly tricky, and there's a high risk of accidentally using future data, which can lead to data leakage. FeatureExpress is designed to nearly eliminate such mistakes. Moreover, I believe that representing data as events is an intuitive approach.
Right now I start my prompts to chatgpt-4 with "dont be lazy". For every question I have it answers that it is a complex problem... GPT-3.5 in the API is more consistent that chatgpt-4. Even with some additional prompts it is making so many mistakes that it takes me multiple tries to get the right output with conversation resets from time to time to start from the previous solution.
I'm pretty surprised more people dont use logit biases to call openai with. Checking if something is either a or b means that the tokens for those letters must be 100 weight which means they will be chosen no matter what and no other character is allowed.
If you're using OAI function calling, you can define a json schema with boolean values. If you have a handful of values, you can use an enum with a list of possible values.
Unless your prompt seriously conflicts with the schema, it's pretty consistent.
Except Paul McCartney is considered as one of the best song writers of all time. There may be many musicians who can do the same on a technical level but even then he is a significant outlier.
> Except Paul McCartney is considered as one of the best song writers of all time.
This is a loaded observation, basically arguing "because all his songs sold well, he must be a significant outlier in songwriting." Similarly, Apple is considered one the best cell phone manufacturers of all time, or McDonald's is considered one of the best burger manufacturers of all time. Popularity does not equate to whether something is any good. And not all of McCartney's songs are the best songs of all time. Some of his notably terrible songs are still popular because he, himself, is famous (because he was a Beatle). One example is the Wing's song Jet, which is popular enough to still get radio play, yet when critically analyzed falls quite short of, say, Hey, Jude.
But my argument is not that Paul McCartney can't write good songs, only that it is ordinary that a musician can recall to perform music they have heard before in the same way most that read something impactful to them can more or less recall it verbatim.
In my opinion this is like criticising Picasso's paintings for not being anything close to as realistic as photos.
Fine, it's a valid opinion to dislike either for those reasons, but for many people it was, and is, art that's exciting in part because of its perfect non-perfection.
I think with lyrics most people don't give them a second thought and don't realise what the person is singing about. Most people can't even tell you which instrument is being used in a part of a song.
This is different to visual art - most people can play spot the difference no problem, but with audio most people are deaf to the details, and the lyrics!
Eg. My Sharona - most people don't realise it includes the lyric "I always get it up for the younger kind"....
