huangfu/InteractionImitation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Optionally save videos of policy evaluations

Browse Source
This commit is contained in:
ebuehrle
2022-03-04 15:58:18 +01:00
parent a9feec4f38
commit 1e70303c57
9 changed files with 95 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
eval_experiments.py
View File

@@ -6,7 +6,7 @@ import json
activations = [torch.nn.Tanh, torch.nn.LeakyReLU]
def main(method:str='expert', folder:str=None, locations=[(0,0)], skip_running=False):
def main(method:str='expert', folder:str=None, locations=[(0,0)], skip_running=False, save_videos:bool=False, videos_folder:str='videos'):
exclude_keys_from_policy_kwargs = {'learning_rate', 'learning_rate_decay', 'clip_ratio', 'iterations_per_epoch', 'option'}
policy_kwargs = {}
@@ -50,7 +50,8 @@ def main(method:str='expert', folder:str=None, locations=[(0,0)], skip_running=F
policy_file=policy_file,
policy_kwargs=policy_kwargs,
env=env,
env_kwargs=env_kwargs)
env_kwargs=env_kwargs,
videos_folder=None if not save_videos else videos_folder)
outfolder = os.path.dirname(outbase)
else:
locstr = 'loc_'+'_'.join([f'r{ro}t{tr}' for (ro,tr) in locations])