andrewzhang505

Andrewzhang505's group workspace

Group: mujoco_halfcheetah

1-1

of 1

Tags

envpool

Author

andrewzhang505

State

Finished

Start time

December 11th, 2022 6:44:10 AM

Runtime

10m 23s

Tracked hours

10m 13s

Run path

andrewzhang505/sample_factory/envpool_halfcheetah4_20221211_064409_800052

Linux-4.19.0-22-cloud-amd64-x86_64-with-glibc2.28

Python version

3.9.13

Git repository

git clone https://github.com/andrewzhang505/sample-factory.git

Git state

git checkout -b "envpool_halfcheetah4_20221211_064409_800052" 0319b736bfad9cf54f8f3ad6ab669193d3ca1d8b

Command

-m sf_examples.envpool.mujoco.train_envpool_mujoco --algo=APPO --env=mujoco_halfcheetah --experiment=envpool_halfcheetah4 --with_wandb=True --wandb_tags envpool --seed=3333

System Hardware

CPU count	8
GPU count	1
GPU type	Tesla T4

W&B CLI Version

0.12.20

Group

mujoco_halfcheetah

Job Type

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 123 keys
- actor_critic_share_weights:
  true
- actor_worker_gpus:[] 0 items
- adam_beta1:
  0.9
- adam_beta2:
  0.999
- adam_eps:
  0.000001
- adaptive_stddev:
  false
- algo:
  "APPO"
- async_rl:
  false
- batch_size:
  2,048
- batched_sampling:
  true
- benchmark:
  false
- ▶
  cli_args:{} 6 keys
- command_line:
  "--algo=APPO --env=mujoco_halfcheetah --experiment=envpool_halfcheetah4 --with_wandb=True --wandb_tags envpool --seed=3333"
- continuous_tanh_scale:
  0
- decoder_mlp_layers:[] 0 items
- decorrelate_envs_on_one_worker:
  true
- decorrelate_experience_max_seconds:
  0
- default_niceness:
  0
- device:
  "gpu"
- encoder_conv_architecture:
  "convnet_simple"
- ▶
  encoder_conv_mlp_layers:[] 1 item
  - 0:
    512
- ▶
  encoder_mlp_layers:[] 3 items
  - 0:
    128
  - 1:
    64
  - 2:
    32
- env:
  "mujoco_halfcheetah"
- env_agents:
  64
- env_frameskip:
  1
- env_framestack:
  1
- env_gpu_actions:
  false
- env_gpu_observations:
  true
- envpool_num_threads:
  8
- envpool_thread_affinity_offset:
  -1
- experiment:
  "envpool_halfcheetah4"
- experiment_summaries_interval:
  15
- exploration_loss:
  "entropy"
- exploration_loss_coeff:
  0
- flush_summaries_interval:
  30
- force_envs_single_thread:
  false
- gae_lambda:
  0.95
- gamma:
  0.99
- git_hash:
  "0319b736bfad9cf54f8f3ad6ab669193d3ca1d8b"
- git_repo_name:
  https://github.com/andrewzhang505/sample-factory.git
- heartbeat_interval:
  20
- heartbeat_reporting_interval:
  180
- help:
  false
- initial_stddev:
  1
- keep_checkpoints:
  2
- kl_loss_coeff:
  0.1
- with_pbt:
  false
- with_vtrace:
  false
- with_wandb:
  true
- worker_num_splits:
  1

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 63 keys
- global_step:
  9,945,088
- len/len:
  1,000
- len/len_max:
  1,000
- len/len_min:
  1,000
- perf/_fps:
  16,589.234375
- perf/_sample_throughput:
  16,821.73046875
- policy_stats/avg_max_raw_reward:
  9.656017303466797
- policy_stats/avg_min_raw_reward:
  -1.169283390045166
- reward/reward:
  6,638.82177734375
- reward/reward_max:
  7,099.32763671875
- reward/reward_min:
  979.3792724609376
- stats/avg_request_count:
  1
- stats/gpu_cache_learner:
  29.36012840270996
- stats/gpu_cache_policy_worker:
  29.36012840270996
- stats/gpu_mem_learner:
  5.620736122131348
- stats/gpu_mem_policy_worker:
  4.085247993469238
- stats/master_process_memory_mb:
  3,148.94140625
- stats/memory_learner:
  3,148.93359375
- stats/memory_policy_worker:
  3,148.94140625
- stats/step_policy:
  0.0014573478838428855
- stats/wait_policy:
  0
- train/action_log_std_max:
  -1.5782458782196045
- train/action_log_std_mean:
  -2.0783376693725586
- train/action_log_std_min:
  -2.809431552886963
- train/action_mean:
  0.056987226009368896
- train/action_mean_max:
  2.0868031978607178
- train/action_mean_min:
  -3.008312463760376
- train/action_stddev_max:
  0.20633670687675476
- train/action_stddev_mean:
  0.13815772533416748
- train/action_stddev_min:
  0.06023922562599182
- train/actual_lr:
  0.00014814814494457096
- train/adam_max_second_moment:
  0.022659189999103543
- train/adv_max:
  4.661752700805664
- train/adv_min:
  -4.448618412017822
- train/adv_std:
  0.9262335300445556
- train/entropy:
  -3.9563956260681152
- train/exploration_loss:
  0
- train/fraction_clipped:
  0.16796875
- train/grad_norm:
  0.9999999403953552
- train/kl_divergence:
  0.01102251186966896
- train/kl_divergence_max:
  0.0731877014040947
- train/kl_loss:
  0.0011022512335330248
- train/loss:
  0.0024389743339270353
- train/lr:
  0.00014814814494457096
- train/max_abs_logprob:
  3.0022132396698
- train/num_sgd_steps:
  40
- train/value_loss:
  0.004367459565401077
- train/version_diff_avg:
  39
- train/version_diff_max:
  39
- train/version_diff_min:
  39

This run produced these artifacts as outputs. Total: 1. Learn more

wandb-history

run-envpool_halfcheetah4_20221211_064409_800052-history:v0