Comment by zamadatix 4 days ago You don't need the previous training material to customize the weights. 3 comments zamadatix Reply sokoloff 3 days ago I don’t need the source code to randomly change bytes in the compiled Linux kernel binary either. 0-_-0 11 hours ago Source code makes it easier to modify a binary, training data doesn't make it easier to modify a trained network. ben_w 3 days ago Fine-tuning weights is easier than retraining the foundation model from scratch with a different corpus.
sokoloff 3 days ago I don’t need the source code to randomly change bytes in the compiled Linux kernel binary either. 0-_-0 11 hours ago Source code makes it easier to modify a binary, training data doesn't make it easier to modify a trained network. ben_w 3 days ago Fine-tuning weights is easier than retraining the foundation model from scratch with a different corpus.
0-_-0 11 hours ago Source code makes it easier to modify a binary, training data doesn't make it easier to modify a trained network.
ben_w 3 days ago Fine-tuning weights is easier than retraining the foundation model from scratch with a different corpus.
I don’t need the source code to randomly change bytes in the compiled Linux kernel binary either.
Source code makes it easier to modify a binary, training data doesn't make it easier to modify a trained network.
Fine-tuning weights is easier than retraining the foundation model from scratch with a different corpus.