[rllib] Expose algorithm parameters and tune policy gradient paramete…

…rs for humanoid (ray-project#753)

* parameters for humanoid

* fix

python/ray/rllib/test.sh

@@ -1,6 +1,7 @@
 #!/bin/bash
 
 python train.py --env Walker2d-v1 --alg PolicyGradient --upload-dir s3://bucketname/
+python train.py --env Humanoid-v1 --alg PolicyGradient --config '{"kl_coeff": 1.0, "num_sgd_iter": 20, "sgd_stepsize": 1e-4, "sgd_batchsize": 32768, "devices": ["/gpu:0", "/gpu:1", "/gpu:2", "/gpu:3"], "tf_session_args": {"device_count": {"GPU": 4}, "log_device_placement": false, "allow_soft_placement": true}, "timesteps_per_batch": 320000, "num_agents": 64}' --upload-dir s3://bucketname/
 python train.py --env PongNoFrameskip-v0 --alg DQN --upload-dir s3://bucketname/
 python train.py --env PongDeterministic-v0 --alg A3C --upload-dir s3://bucketname/
 python train.py --env Humanoid-v1 --alg EvolutionStrategies --upload-dir s3://bucketname/

python/ray/rllib/train.py

@@ -18,6 +18,7 @@
     description=("Train a reinforcement learning agent."))
 parser.add_argument("--env", required=True, type=str)
 parser.add_argument("--alg", required=True, type=str)
+parser.add_argument("--config", default="{}", type=str)
 parser.add_argument("--upload-dir", default="file:///tmp/ray", type=str)
 
 
@@ -28,17 +29,25 @@
 
     env_name = args.env
     if args.alg == "PolicyGradient":
+        config = pg.DEFAULT_CONFIG.copy()
+        config.update(json.loads(args.config))
         alg = pg.PolicyGradient(
-            env_name, pg.DEFAULT_CONFIG, upload_dir=args.upload_dir)
+            env_name, config, upload_dir=args.upload_dir)
     elif args.alg == "EvolutionStrategies":
+        config = es.DEFAULT_CONFIG.copy()
+        config.update(json.loads(args.config))
         alg = es.EvolutionStrategies(
-            env_name, es.DEFAULT_CONFIG, upload_dir=args.upload_dir)
+            env_name, config, upload_dir=args.upload_dir)
     elif args.alg == "DQN":
+        config = dqn.DEFAULT_CONFIG.copy()
+        config.update(json.loads(args.config))
         alg = dqn.DQN(
-            env_name, dqn.DEFAULT_CONFIG, upload_dir=args.upload_dir)
+            env_name, config, upload_dir=args.upload_dir)
     elif args.alg == "A3C":
+        config = a3c.DEFAULT_CONFIG.copy()
+        config.update(json.loads(args.config))
         alg = a3c.A3C(
-            env_name, a3c.DEFAULT_CONFIG, upload_dir=args.upload_dir)
+            env_name, config, upload_dir=args.upload_dir)
     else:
         assert False, ("Unknown algorithm, check --alg argument. Valid "
                        "choices are PolicyGradientPolicyGradient, "