A question about finetuning on RSICD

[mcx-mcx]

Thank you very much for your outstanding work！
I have a question that I haven't quite understood. When fine-tuning your RS5M model on RSICD or RSITMD using the methods outlined in the paper (infoNCE, lr=1e-6), I did not achieve the expected performance. Taking RSICD as an example, the paper and the weights you provided for RS5M RET-2 version result in an accuracy around 38, but when I fine-tuned using my own RS5M VitB32 version, the result was around 34. Could you provide more details on fine-tuning RET-2 or RSICD so that I can better replicate the process? Thank you very much.

[zilunzhang]

Thank you very much for your outstanding work！ I have a question that I haven't quite understood. When fine-tuning your RS5M model on RSICD or RSITMD using the methods outlined in the paper (infoNCE, lr=1e-6), I did not achieve the expected performance. Taking RSICD as an example, the paper and the weights you provided for RS5M RET-2 version result in an accuracy around 38, but when I fine-tuned using my own RS5M VitB32 version, the result was around 34. Could you provide more details on fine-tuning RET-2 or RSICD so that I can better replicate the process? Thank you very much.

Hi

We used the ITRA codebase to tune the model.

The RET-2 data was made by: ChenDelong1999/RemoteCLIP#13 (comment)

Hyperparams for RET-2 model:

--lr: 5e-06
--weight_decay: 0.5
--batch_size: 600
--epochs: 7
--warmup 100
--max_grad_norm: 50.0

Best,

Zilun

[mcx-mcx]

Thank you very much!

[lemyx]

Thanks for your great work!

I wanna reproduce the following results, could you share the hyperparameters related to lr, weight_decay, bs, epochs, warmup and max_grad_norm?

[zilunzhang]

Thanks for your great work!

I wanna reproduce the following results, could you share the hyperparameters related to lr, weight_decay, bs, epochs, warmup and max_grad_norm?

Sure. Could you leave us an email address?

[r-deo]

@zilunzhang I was trying the reproduce FT on RS5M. But, I was not able to reproduce. Would it be possible to share the hyper parameter and training related modifications?

Thanks