pusht数据集配置，优化后测试结果

config/eval/pusht.yaml

@@ -19,6 +19,7 @@ dataset:
 
 seed: 42
 policy: random # ckpt name or random
+inference_precision: fp16
 
 plan_config:
   horizon: 5
@@ -45,4 +46,4 @@ eval:
           value: goal_state
           
 output:
-  filename: pusht_results.txt
+  filename: pusht_results.txt

pusht_results.txt

@@ -0,0 +1,439 @@
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+
+==== RESULTS ====
+metrics: {'success_rate': 92.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True, False,
+        True, False,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 68.59998798370361 seconds
+inference_precision: fp32
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+
+==== RESULTS ====
+metrics: {'success_rate': 92.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True, False,
+        True, False,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 45.66186594963074 seconds
+inference_precision: fp32
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+multi_gpu:
+  enabled: true
+  devices:
+  - 0
+  - 1
+
+==== RESULTS ====
+metrics: {'success_rate': 92.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True, False,  True,
+        True, False,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True, False,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 69.21508717536926 seconds
+inference_precision: fp32
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+inference_precision: fp16
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+multi_gpu:
+  enabled: true
+  devices:
+  - 0
+  - 1
+
+==== RESULTS ====
+metrics: {'success_rate': 94.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True, False,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True, False,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 37.82094120979309 seconds
+inference_precision: fp16
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+inference_precision: fp16
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+multi_gpu:
+  enabled: true
+  devices:
+  - 0
+  - 1
+
+==== RESULTS ====
+metrics: {'success_rate': 94.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True, False,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True, False,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 37.55708169937134 seconds
+inference_precision: fp16
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+inference_precision: fp16
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+
+==== RESULTS ====
+metrics: {'success_rate': 94.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True, False,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True, False,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 29.592512607574463 seconds
+inference_precision: fp16
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead
+
+==== CONFIG ====
+cache_dir: /mnt/ASC1637/stablewm
+solver:
+  _target_: stable_worldmodel.solver.CEMSolver
+  model: ???
+  batch_size: 8
+  num_samples: 300
+  var_scale: 1.0
+  n_steps: 30
+  topk: 30
+  device: cuda
+  seed: ${seed}
+world:
+  env_name: swm/PushT-v1
+  num_envs: ${eval.num_eval}
+  max_episode_steps: ???
+  history_size: 1
+  frame_skip: 1
+dataset:
+  stats: ${eval.dataset_name}
+  keys_to_cache:
+  - action
+  - proprio
+  - state
+seed: 42
+policy: pusht/lewm
+inference_precision: fp16
+plan_config:
+  horizon: 5
+  receding_horizon: 5
+  action_block: 5
+eval:
+  num_eval: 50
+  goal_offset_steps: 25
+  eval_budget: 50
+  img_size: 224
+  dataset_name: pusht_expert_train
+  callables:
+  - method: _set_state
+    args:
+      state:
+        value: state
+  - method: _set_goal_state
+    args:
+      goal_state:
+        value: goal_state
+output:
+  filename: pusht_results.txt
+
+==== RESULTS ====
+metrics: {'success_rate': 94.0, 'episode_successes': array([ True,  True,  True,  True,  True, False,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True, False,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True,  True,  True,  True,
+        True,  True,  True,  True,  True,  True, False,  True,  True,
+        True,  True,  True,  True,  True]), 'seeds': None}
+evaluation_time: 29.783243894577026 seconds
+inference_precision: fp16
+inference_compile_target: predictor
+inference_compile_mode: reduce-overhead