wassname/ray
1
0
Fork 0
mirror of https://github.com/wassname/ray.git synced 2026-07-10 01:39:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

For PPO, rename num_agents -> num_workers. (#882)

Browse Source
This commit is contained in:
Robert Nishihara
2017-08-28 23:11:06 -07:00
committed by Philipp Moritz
parent 1afc487baf
commit e1831792f8
3 changed files with 9 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/ray/rllib/policy_gradient/policy_gradient.py
+2 -2
View File
@@ -60,7 +60,7 @@ DEFAULT_CONFIG = {
# number of steps is obtained
"min_steps_per_task": 1000,
# Number of actors used to collect the rollouts
"num_agents": 5,
"num_workers": 5,
# Dump TensorFlow timeline after this many SGD minibatches
"full_trace_nth_sgd_batch": -1,
# Whether to profile data loading
@@ -92,7 +92,7 @@ class PolicyGradient(Algorithm):
self.agents = [
RemoteAgent.remote(
self.env_name, 1, self.config, self.logdir, True)
for _ in range(self.config["num_agents"])]
for _ in range(self.config["num_workers"])]
self.start_time = time.time()
if self.config["write_logs"]:
self.file_writer = tf.summary.FileWriter(
python/ray/rllib/test/test.sh
+5 -5
View File
@@ -1,14 +1,14 @@
#!/bin/bash
python train.py --env Hopper-v1 --config '{"gamma": 0.995, "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": 160000, "num_agents": 64}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Hopper-v1 --config '{"gamma": 0.995, "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": 160000, "num_workers": 64}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env CartPole-v1 --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": 160000, "num_agents": 64}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env CartPole-v1 --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": 160000, "num_workers": 64}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Walker2d-v1 --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}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Walker2d-v1 --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_workers": 64}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Humanoid-v1 --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, "model": {"free_log_std": true}, "use_gae": false}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Humanoid-v1 --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_workers": 64, "model": {"free_log_std": true}, "use_gae": false}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Humanoid-v1 --config '{"lambda": 0.95, "clip_param": 0.2, "kl_coeff": 1.0, "num_sgd_iter": 20, "sgd_stepsize": 1e-4, "sgd_batchsize": 32768, "horizon": 5000, "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, "model": {"free_log_std": true}, "write_logs": false}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env Humanoid-v1 --config '{"lambda": 0.95, "clip_param": 0.2, "kl_coeff": 1.0, "num_sgd_iter": 20, "sgd_stepsize": 1e-4, "sgd_batchsize": 32768, "horizon": 5000, "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_workers": 64, "model": {"free_log_std": true}, "write_logs": false}' --alg PolicyGradient --upload-dir s3://bucketname/
python train.py --env PongNoFrameskip-v0 --alg DQN --upload-dir s3://bucketname/