How do I decide which AI/ML model is best suited for the problem I want to solve?
Solid question, I’ll try to give the most detail I can without overwhelming you:
The short answer is: Look for models specific for your use case. Like Whisper for transcription, or Voxtral for TTS, etc.
The long answer is, define your use case. Then when you go to the models page, you should filter by “tasks”:
This filter gives you a massive list of things that models can be tagged to specifically handle. Usually you can find something there that matches.
You’ll also have to keep in mind how you’ll be running the model, its size, the hardware needed, etc.
Best of luck!
Here’s a page for an introduction to AI, and happens to list some AI models good at coding. Introduction to AI
There are AI’s ranked by test results at <https://llm-stats.com/>. One column is “SWE-BENCH” which is for code changes. The page above also has other sites for ranking AI, some are ranked by user votes.
Huggingface.co has categories for AI. Check there also.
Well you first need to know what hardware you will run your models on for local usage. Onec you know what your hardware will support then you can look into models that fit your use case.
