aichat VS unstructured-inference

thanks for sharing this, I have been using aichat (https://github.com/sigoden/aichat) and shell_gpt for a while. Let's see how it works.

use LLMs (gpt-4-vision or LLaVA) with aichat

`aichat -f tmp/test.png -- output only text in the image`

https://github.com/sigoden/aichat

Tools required: https://github.com/yt-dlp/yt-dlp https://github.com/sigoden/aichat

Yes, I'm using for language learning. Usually via tool `https://github.com/sigoden/aichat` with different roles that I've created. For example for grammar ChatGPT generates bunch of examples for specific grammar rules, which I memorize.

I've been enjoying using aichat for command-line chatgpt access https://github.com/sigoden/aichat (it supports both an interactive repl mode and you can invoke it with your query directly from the command line like the helper used in this article).

https://github.com/sigoden/aichat Interactive or run it in one line Temporarily turn on multiple-line mode when you need

Tesseract is widely known to be "meh" at this point.

If you look at RAG frameworks as one example they'll typically use/support a variety of implementations. Tesseract is almost always supported but it's rarely ideal with projects like Unstructured[0] and DocTR[1] being preferred. By leveraging more-or-less SOTA vision models[2][3] they embarrass Tesseract.

I haven't compared them to the Apple Vision framework but they're absolutely better than Tesseract and potentially even Apple Vision.

[0] - https://github.com/Unstructured-IO/unstructured-inference

[1] - https://github.com/mindee/doctr

[2] - https://github.com/mindee/doctr#models-architectures

[3] - https://github.com/Unstructured-IO/unstructured-inference#mo...