Closed fcakyon closed 1 year ago
Hi @fcakyon . The license does not limit any fork. the code itself can be used under the Apache 2.0 license terms. You can train the model from scratch without using the pre-trained weight under the Apache 2.0 license terms.
Only if you are using the pre-trained weights you are bounded to the non-commercial license. If you fine-tune your model based on the pte-trained weights then you are bounded to the non-commercial license.
If we use the pre-trained models for knowledge distillation to pre-train the Yolo-NAS from scratch, we're not strictly using the weights. In this case, can we get round the licence ?
Hello,
As far as I understood, YOLONAS license prevents the commercial use of any forks of the repo with custom architectural changes. It also prevents the commercial use of any custom finetuned model. Am I right on that?
https://github.com/Deci-AI/super-gradients/blob/1c67ce369b53daf162a91faacb24f3501e7b84ad/LICENSE.YOLONAS.md