George Hotz 49bcfec383 0s in the action space (#2070) ...

* 0s in the action space

* simpler

* skip duplicate actions

2023-10-14 11:22:48 -07:00

..

generate_dataset.sh

op logger + replay (#2021)

2023-10-08 15:10:18 -07:00

get_action_space.py

0s in the action space (#2070)

2023-10-14 11:22:48 -07:00

helpers.py

train value net, improve API, add BCE (#2047)

2023-10-12 07:56:38 -07:00

pretrain_policynet.py

train value net, improve API, add BCE (#2047)

2023-10-12 07:56:38 -07:00

pretrain_valuenet.py

train value net, improve API, add BCE (#2047)

2023-10-12 07:56:38 -07:00

rl.py

train value net, improve API, add BCE (#2047)

2023-10-12 07:56:38 -07:00

test_net.py

train value net, improve API, add BCE (#2047)

2023-10-12 07:56:38 -07:00

test_time_linearizer.py

with unroll, the action space goes from 161 -> 127 (#2060)

2023-10-12 20:52:23 -07:00