slm_lab/spec/experimental/sarsa/sarsa_cartpole.json
{
"sarsa_epsilon_greedy_cartpole": {
"agent": [{
"name": "SARSA",
"algorithm": {
"name": "SARSA",
"action_pdtype": "Argmax",
"action_policy": "epsilon_greedy",
"explore_var_spec": {
"name": "linear_decay",
"start_val": 1.0,
"end_val": 0.05,
"start_step": 0,
"end_step": 10000
},
"gamma": 0.99,
"training_frequency": 32
},
"memory": {
"name": "OnPolicyBatchReplay"
},
"net": {
"type": "MLPNet",
"hid_layers": [64],
"hid_layers_activation": "selu",
"clip_grad_val": 0.5,
"loss_spec": {
"name": "MSELoss"
},
"optim_spec": {
"name": "RMSprop",
"lr": 0.01
},
"lr_scheduler_spec": null
}
}],
"env": [{
"name": "CartPole-v0",
"max_t": null,
"max_frame": 80000
}],
"body": {
"product": "outer",
"num": 1
},
"meta": {
"distributed": false,
"eval_frequency": 2000,
"max_trial": 1,
"max_session": 1
}
},
"sarsa_boltzmann_cartpole": {
"agent": [{
"name": "SARSA",
"algorithm": {
"name": "SARSA",
"action_pdtype": "Argmax",
"action_policy": "boltzmann",
"explore_var_spec": {
"name": "linear_decay",
"start_val": 3.0,
"end_val": 1.0,
"start_step": 0,
"end_step": 10000
},
"gamma": 0.99,
"training_frequency": 32
},
"memory": {
"name": "OnPolicyBatchReplay"
},
"net": {
"type": "MLPNet",
"hid_layers": [64],
"hid_layers_activation": "selu",
"clip_grad_val": 0.5,
"loss_spec": {
"name": "MSELoss"
},
"optim_spec": {
"name": "RMSprop",
"lr": 0.01
},
"lr_scheduler_spec": null
}
}],
"env": [{
"name": "CartPole-v0",
"max_t": null,
"max_frame": 80000
}],
"body": {
"product": "outer",
"num": 1
},
"meta": {
"distributed": false,
"eval_frequency": 2000,
"max_trial": 1,
"max_session": 1
}
}
}