ml-model-watermarking
ml-model-watermarking copied to clipboard
Protect your machine learning models easily and securely with watermarking 🔑
I'm trying to run code exaple ------code-------- >>> from mlmodelwatermarking.markface import TrainerWM >>> trainer = TrainerWM(model=your_model) >>> ownership = trainer.watermark() >>> watermarked_model = trainer.get_model() ------------------- It shows an error: ModuleNotFoundError:...
i have runed tweet_sentiment_analysis, but the result is :assert verification['is_stolen'] is True AssertionError, can you help me with it ?
If you look into the examples of hugging face(markface). `from mlmodelwatermarking.markface import Trainer as TrainerWM` There is no Trainer inside mlmodelwatermarking.markface or am I missing something?
The verify function in case of suspected models have a bug. i.e The length of predictions_reference list will not be equal to the length of pred_suspect (line 526 in marktorch.py)...