Tianshou/tianshou/__init__.py

from tianshou import data, env, utils, policy, trainer, exploration

# pre-compile some common-type function-call to produce the correct benchmark
# result: https://github.com/thu-ml/tianshou/pull/193#discussion_r480536371
utils.pre_compile()


__version__ = '0.2.6'

__all__ = [
    'env',
    'data',
    'utils',
    'policy',
    'trainer',
    'exploration',
]
Numba acceleration (#193) Training FPS improvement (base commit is 94bfb32): test_pdqn: 1660 (without numba) -> 1930 discrete/test_ppo: 5100 -> 5170 since nstep has little impact on overall performance, the unit test result is: GAE: 4.1s -> 0.057s nstep: 0.3s -> 0.15s (little improvement) Others: - fix a bug in ttt set_eps - keep only sumtree in segment tree implementation - dirty fix for asyncVenv check_id test 2020-09-02 13:03:32 +08:00			`from tianshou import data, env, utils, policy, trainer, exploration`

			`# pre-compile some common-type function-call to produce the correct benchmark`
			`# result: https://github.com/thu-ml/tianshou/pull/193#discussion_r480536371`
			`utils.pre_compile()`
init 2020-03-09 11:38:04 +08:00
optimize training procedure and improve code coverage (#189) 1. add policy.eval() in all test scripts' "watch performance" 2. remove dict return support for collector preprocess_fn 3. add `__contains__` and `pop` in batch: `key in batch`, `batch.pop(key, deft)` 4. exact n_episode for a list of n_episode limitation and save fake data in cache_buffer when self.buffer is None (#184) 5. fix tensorboard logging: h-axis stands for env step instead of gradient step; add test results into tensorboard 6. add test_returns (both GAE and nstep) 7. change the type-checking order in batch.py and converter.py in order to meet the most often case first 8. fix shape inconsistency for torch.Tensor in replay buffer 9. remove `**kwargs` in ReplayBuffer 10. remove default value in batch.split() and add merge_last argument (#185) 11. improve nstep efficiency 12. add max_batchsize in onpolicy algorithms 13. potential bugfix for subproc.wait 14. fix RecurrentActorProb 15. improve the code-coverage (from 90% to 95%) and remove the dead code 16. fix some incorrect type annotation The above improvement also increases the training FPS: on my computer, the previous version is only ~1800 FPS and after that, it can reach ~2050 (faster than v0.2.4.post1). 2020-08-27 12:15:18 +08:00
code refactor for venv (#179) - Refacor code to remove duplicate code - Enable async simulation for all vector envs - Remove `collector.close` and rename `VectorEnv` to `DummyVectorEnv` The abstraction of vector env changed. Prior to this pr, each vector env is almost independent. After this pr, each env is wrapped into a worker, and vector envs differ with their worker type. In fact, users can just use `BaseVectorEnv` with different workers, I keep `SubprocVectorEnv`, `ShmemVectorEnv` for backward compatibility. Co-authored-by: n+e <463003665@qq.com> Co-authored-by: magicly <magicly007@gmail.com> 2020-08-19 15:00:24 +08:00			`__version__ = '0.2.6'`
Numba acceleration (#193) Training FPS improvement (base commit is 94bfb32): test_pdqn: 1660 (without numba) -> 1930 discrete/test_ppo: 5100 -> 5170 since nstep has little impact on overall performance, the unit test result is: GAE: 4.1s -> 0.057s nstep: 0.3s -> 0.15s (little improvement) Others: - fix a bug in ttt set_eps - keep only sumtree in segment tree implementation - dirty fix for asyncVenv check_id test 2020-09-02 13:03:32 +08:00
maybe finished collector? 2020-03-13 17:49:22 +08:00			`__all__ = [`
			`'env',`
clear 2020-03-13 21:47:17 +08:00			`'data',`
maybe finished collector? 2020-03-13 17:49:22 +08:00			`'utils',`
ddpg 2020-03-18 21:45:41 +08:00			`'policy',`
add trainer 2020-03-19 17:23:46 +08:00			`'trainer',`
ddpg 2020-03-18 21:45:41 +08:00			`'exploration',`
maybe finished collector? 2020-03-13 17:49:22 +08:00			`]`