forked from ChengshuLi/agents
-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathintegration_tests.txt
More file actions
11 lines (9 loc) · 917 Bytes
/
integration_tests.txt
File metadata and controls
11 lines (9 loc) · 917 Bytes
1
2
3
4
5
6
7
8
9
10
11
# This is the set of integration tests to run upon release.
# setup.py automatically applies a unique root_dir to these binaries.
# agents/dqn/examples/v1
tf_agents/agents/dqn/examples/v1/train_eval_atari.py --game_name=Pong --num_iterations=1 --initial_collect_steps=4 --replay_buffer_capacity=10000 --train_steps_per_iteration=2000 --eval_steps_per_iteration=2000 -logtostderr
tf_agents/agents/dqn/examples/v1/oog_train_eval.py --num_iterations=5 --logtostderr
tf_agents/agents/dqn/examples/v1/train_eval_gym.py --num_iterations=2500 --logtostderr
tf_agents/agents/dqn/examples/v1/train_eval_gym.py --num_iterations=2500 --logtostderr --use_ddqn
# agents/categorical_dqn/examples
tf_agents/agents/categorical_dqn/examples/train_eval_atari.py --game_name=Pong --num_iterations=1 --initial_collect_steps=4 --replay_buffer_capacity=10000 --train_steps_per_iteration=2000 --eval_steps_per_iteration=2000 -logtostderr