Open soi opened 3 years ago
Maybe it now works with DanQ. One could also try SWATS to finetune with SGD, but this seems to be only available for pytorch.
https://www.tensorflow.org/api_docs/python/tf/keras/optimizers/schedules/ExponentialDecay
Maybe it now works with DanQ. One could also try SWATS to finetune with SGD, but this seems to be only available for pytorch.