Closed nealmahajan closed 1 year ago
When I run:
time_limit = 60*60*8 # at most 1 hour predictor = ImagePredictor() hyperparameters = {'batch_size': 10} predictor.fit(train, hyperparameters=hyperparameters, hyperparameter_tune_kwargs={'num_trials': 2, 'searcher': 'random'}, presets='good_quality_fast_inference', time_limit=time_limit) predictor.fit_summary()
the run terminates with KeyError: 'estimator'. However, I am able to run the object detection package on my gpu without issue. Do you know what the issue may be?
Additionally, I don't get this error when I have 4 labels in my dataset, but do when I have only 2.
Fixed in v0.7: recommendation to use autogluon.multimodal for vision tasks
When I run:
the run terminates with KeyError: 'estimator'. However, I am able to run the object detection package on my gpu without issue. Do you know what the issue may be?