FINE-TUNING with Custom Dataset

[SyedMohammedSameer]

Search before asking

• I have searched the Multimodal Maestro issues and found no similar feature requests.

Question

Firstly, Great work and congratulations on creating Maestro! I'm really impressed with its capabilities for fine-tuning Vision LLMs.
I'm particularly interested in using Maestro with a custom dataset. I have some images for which I'm annotating manually.
I specifically have doubts on creating the JSONL and how to use it with Maestro.
Could you share some insights or steps on how to get started with that?
Any guidance would be much appreciated!

Thanks in advance!

Additional

No response

[SkalskiP]

Hi @SyedMohammedSameer 👋🏻 Thank you so much for your kind words. I added a docs page this morning diving deeper into the JSONL format. I hope it will be helpful. I'm closing the issue. Feel free to reopen it if you still have questions.