Any Stack Efficientnet-B0 Test Results v0.1

Here are the first V0.1 results of CoSTAR Visual Stacking.
This test run is configured so that any stack is considered a success.
Logged data and the model at the end is attached.
Click on the image below for the complete test run video:

V0.1 First stacking simulation only numerical summary test results are as follows:

Testing iteration: 226
Change detected: True (value: 659)
Primitive confidence scores: 1.865717 (push), 2.080815 (grasp), 2.263923 (place)
Action: place at (8, 50, 107)
Current reward: 0.375000
Future reward: 2.283270
Expected reward: 0.375000 + 0.500000 x 2.283270 = 1.516635
Executing: place at (-0.510000, -0.124000, 0.154729)
Training loss: 0.002941
gripper position: 0.018331587314605713
gripper position: 0.01809576153755188
current_position: [-0.52042723 -0.12329237  0.18156546]
current_obj_z_location: 0.21156546354293823
goal_position: 0.21472921353236296 goal_position_margin: 0.31472921353236294
has_moved: True near_goal: True place_success: True
check_stack() current detected stack height: 4
check_stack() stack_height: 4 stack matches current goal: True partial_stack_success: True Does the code think a reset is needed: False
STACK:  trial: 51 actions/partial: 2.986842105263158  actions/full stack: 14.1875 (lower is better)  Grasp Count: 118, grasp success rate: 0.847457627118644 place_on_stack_rate: 0.76 place_attempts: 100  partial_stack_successes: 76  stack_successes: 16 trial_success_rate: 0.3137254901960784 stack goal: [1 0 3 2]
Time elapsed: 9.919075
Trainer iteration: 227.000000

Below is an example visualization of the Q function for each action in the order grasp, place, push, the final action actually taken was place:

Command line command to collect this data:

export CUDA_VISIBLE_DEVICES="0" && python3 main.py --is_sim --obj_mesh_dir 'objects/blocks' --num_obj 4  --push_rewards --experience_replay --explore_rate_decay --place --load_snapshot --snapshot_file '/home/costar/src/costar_visual_stacking/logs/2019-08-15.12:38:20/models/snapshot.reinforcement-best-stack-rate.pth' --random_seed 1238 --is_testing --save_visualizations