Fine Tune Qwen2 VL Model using Llama Factory

[@user4-j1w]

Helped me in my fine tuning on image dataset

[@excalibur2889]

How to load your own dataset?

[@sneharoy3566]

awaited video.. Thank you

[@AIAnytime]

Hope you enjoyed it!

[@atultiwari88]

thank you so much

[@AIAnytime]

You're welcome!

[@arjunreddy2556]

Should the custom dataset be in sharegpt format always for fine-tuning? If not, can we use another JSON format?

[@shortleg18]

How to use different datasets and inferencing???

[@MayurKuchhadiya-f9b]

How can we change it rank(r) and alpha value?

[@satheeshthangaraj5614]

Hi, pls let us know how to use the custom dataset from local folder or dataset from huggingface, which parameter we should change.

[@IsmailIfakir]

multimodal llm can fine-tuning for sentiment analysis

[@explorer945]

can you change your camera settings to mirror your video so that we can see the text correctly in the background (manchester united and other drawings on the board)

[@AIAnytime]

Sure will do that.....

[@azizahtas]

Awesome video. I am the first one to comment. 🎉🎉

[@AIAnytime]

Thanks

[@nguyenanhnguyen7658]

LlaMA-Factory is buggy. You can NOT finetune with vision tower open.

[@cyberthugFi]

Hey, thanks for. the video, you quite a smart guy, my question is about the Vision features of Qwen2 VL, can i wire my camera to make it able to my model to see?

[@NatnaelTamirat-b4p]

i like you videos but the flag behind you uhhh

[@satvik4225]

Can you please let me know how I could do inference, I had done training with the radio ui, have model saved in /LLaMA-Factory/saves/Qwen2VL-2B-Chat/lora/train_2024-09-13-12-36-15

[@AlphaGodz]

Thank you So much