HPARAMS = {
"horizon": 256,
"num_epochs": 10,
"batch_size": 15,
"policy_clip": 0.1,
"discount_factor": 0.99,
"GAE_lambda": 0.95,
"entropy_coeff": 0.01,
"value_coeff": 0.5,
"learning_rate": 0.0003,
}