slm_lab/spec/benchmark/sarsa/sarsa_videopinball.json
{
"sarsa_epsilon_greedy_videopinball": {
"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": 1000,
"end_step": 100000
},
"gamma": 0.99,
"training_frequency": 32
},
"memory": {
"name": "OnPolicyBatchReplay",
"batch_size": 64
},
"net": {
"type": "ConvNet",
"conv_hid_layers": [
[32, 8, 4, 0, 1],
[64, 4, 2, 0, 1],
[32, 3, 1, 0, 1]
],
"fc_hid_layers": [512],
"hid_layers_activation": "relu",
"init_fn": "orthogonal_",
"normalize": true,
"batch_norm": false,
"use_same_optim": false,
"clip_grad_val": 0.5,
"loss_spec": {
"name": "MSELoss"
},
"optim_spec": {
"name": "RMSprop",
"lr": 0.01
},
"lr_scheduler_spec": null,
"gpu": true
}
}],
"env": [{
"name": "VideoPinball-v0",
"frame_op": "concat",
"frame_op_len": 4,
"reward_scale": "sign",
"num_envs": 16,
"max_t": null,
"max_frame": 850000
}],
"body": {
"product": "outer",
"num": 1
},
"meta": {
"distributed": false,
"eval_frequency": 2000,
"max_trial": 1,
"max_session": 1
}
},
"sarsa_boltzmann_videopinball": {
"agent": [{
"name": "SARSA",
"algorithm": {
"name": "SARSA",
"action_pdtype": "Argmax",
"action_policy": "boltzmann",
"explore_var_spec": {
"name": "linear_decay",
"start_val": 1.0,
"end_val": 0.05,
"start_step": 1000,
"end_step": 100000
},
"gamma": 0.99,
"training_frequency": 32
},
"memory": {
"name": "OnPolicyBatchReplay",
"batch_size": 64
},
"net": {
"type": "ConvNet",
"conv_hid_layers": [
[32, 8, 4, 0, 1],
[64, 4, 2, 0, 1],
[32, 3, 1, 0, 1]
],
"fc_hid_layers": [512],
"hid_layers_activation": "relu",
"init_fn": "orthogonal_",
"normalize": true,
"batch_norm": false,
"use_same_optim": false,
"clip_grad_val": 0.5,
"loss_spec": {
"name": "MSELoss"
},
"optim_spec": {
"name": "RMSprop",
"lr": 0.01
},
"lr_scheduler_spec": null,
"gpu": true
}
}],
"env": [{
"name": "VideoPinball-v0",
"frame_op": "concat",
"frame_op_len": 4,
"reward_scale": "sign",
"num_envs": 16,
"max_t": null,
"max_frame": 850000
}],
"body": {
"product": "outer",
"num": 1
},
"meta": {
"distributed": false,
"eval_frequency": 2000,
"max_trial": 1,
"max_session": 1
}
}
}