Conversion of inputs to Tensorflow problem in RL training of Models Q and R

Course Name: Deep Reinforcement Learning in Trading, Section No: 18, Unit No: 1, Unit type: Video

I'm experiencing difficulties with the experience replay code. TensorFlow appears to not receive the data passed as input for model R  inputs, targets = exp_replay.process()



 targets[i] = modelR.predict(state_t)[0]

 t = convert_to_eager_tensor(value, ctx, dtype)

return ops.EagerTensor(value, ctx.device_name, dtype)



ValueError: Failed to convert a NumPy array to a Tensor (Unsupported object type float).

Hi Diego,



The EagerTensor code seems to be new here. Could you please share the code (.ipynb + .py files) and the data files you are working on so we can help debug this issue?



You can send the files over e-mail at quantra@quantinsti.com



Thanks!