It's a small patch to Apple's #OpenSource MLX machine learning framework to fix a minor nit that annoyed me, but I was happy to have my first PR accepted against a Python project. (github.com/ml-explore/mlx-exam…).

Still happily doing #Perl, but #AI is #Python all the way.

Make it easier to know in which file we have bad JSON data by Ovid · Pull Request #458 · ml-explore/mlx-examples

I was using lora.py to train Mistral and accidentally used corrupted JSONL files. My traceback looked like this: Traceback (most recent call last): File "/Users/ovid/projects/llms/mlx-examples/l...

^GitHub