Split CompilerEnv.step() into two methods for singular or lists of actions (take 2) #627
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sogartar
force-pushed
the
feature/deprecate-action-list
branch
from
d0d464a to
6acd592
Compare
sogartar
changed the title
sogartar
changed the title
facebook-github-bot
added
the
CLA Signed
Codecov Report
@@ Coverage Diff @@
## development #627 +/- ##
===============================================
- Coverage 88.61% 79.76% -8.85%
===============================================
Files 115 115
Lines 7019 7058 +39
===============================================
- Hits 6220 5630 -590
- Misses 799 1428 +629
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## development #627 +/- ##
===============================================
- Coverage 88.75% 88.23% -0.53%
===============================================
Files 115 115
Lines 7019 7061 +42
===============================================
Hits 6230 6230
- Misses 789 831 +42
Continue to review full report at Codecov.
|
|
@ChrisCummins, I think I fixed the tests and examples. It should be ready for a review. |
ChrisCummins
approved these changes
Mar 17, 2022
There was a problem hiding this comment.
Hi @sogartar, thanks for pushing ahead with this. I think overloading multistep() makes total sense.
A few small changes requested, otherwise, LGTM!
compiler_gym/wrappers/core.py
Outdated
Comment on lines
129
to
138
| # Undo the episode_reward update and reapply it once we have transformed | ||
| # the reward. | ||
| # | ||
| # TODO(cummins): Refactor step() so that we don't have to do this | ||
| # recalculation of episode_reward, as this is prone to errors if, say, | ||
| # the base reward returns NaN or an invalid type. | ||
| if reward is not None and self.episode_reward is not None: | ||
| self.unwrapped.episode_reward -= reward | ||
| reward = self.reward(reward) | ||
| self.unwrapped.episode_reward += reward |
There was a problem hiding this comment.
I think this workaround needs to be included, because multistep() runs the update to self.episode_reward.
There was a problem hiding this comment.
There should really be a unit test to cover this (not saying you need to add it 🙂 , just note to self)
Comment on lines
+35
to
+36
| observation_spaces=None, | ||
| reward_spaces=None, |
There was a problem hiding this comment.
This also needs to accept observations and rewards arguments for backwards compatibility.
| @@ -5,6 +5,8 @@ | |||
|
|
|||
| cg_add_all_subdirs() | |||
|
|
|||
| return() | |||
There was a problem hiding this comment.
Is this meant to be in here?
tests/wrappers/core_wrappers_test.py
Outdated
| @@ -264,11 +272,9 @@ def reward(self, reward): | |||
| env.reset() | |||
| _, reward, _, _ = env.step(0) | |||
| assert reward == -5 | |||
| assert env.episode_reward == -5 | |||
There was a problem hiding this comment.
I think these two deleted lines can be restored now that you're overloading multistep() rather than raw_step()
|
Heads up that |
CompilerEnv.step() currently accepts two types for the "action" argument: a scalar action, or an iterable of actions. This kind of type overloading does not work for list types. This adds a new method, CompilerEnv.multistep(), that explicitly takes takes an iterable sequence of actions. If you want to run multiple actions in a single step, call this new method. Calling CompilerEnv.step() with a list of actions still works, though with a deprecation warning. In the v0.2.4 release support for lists of actions in CompilerEnv.step() will be removed. Fixes #610.
This makes the following changes: - Changes env.step() `action` to accept only a single action, with a deprecation warning if a list of actions are provided. - Renames env.step() `observations` to `observation_spaces`. The old parameter name is still accepted with a deprecation warning. - Renames env.step() `rewards` to `reward_spaces`. The old parameter name is still accepted with a deprecation warning.
sogartar
force-pushed
the
feature/deprecate-action-list
branch
from
4360737 to
8cd9679
Compare
|
@ChrisCummins, I addressed your comments. If your OK with them I think it is ready to merge after the CI tests have passed. |
This was referenced Mar 17, 2022
|
Brill, thanks! Merging 🎉 |
ChrisCummins
added a commit
that referenced
this pull request
May 24, 2022
This release adds a new compiler environment, new APIs, and a suite of backend improvements to improve the flexibility of CompilerGym environments. Many thanks to code contributors: @sogartar, @KyleHerndon, @SoumyajitKarmakar, @uduse, and @anthony0727! Highlights of this release include: - [mlir] Began work on a new environment for matrix multiplication using MLIR ([#652](#652), thanks @KyleHerndon and @sogartar!). Note this environment is not yet included in the pypi package and must be [compiled from source](https://github.com/facebookresearch/CompilerGym/blob/development/INSTALL.md#building-from-source-with-cmake). - [llvm] Added a new `env.benchmark_from_clang_invocation()` method ([#577](#577)) that can be used for constructing LLVM environment automatically from C/C++ compiler invocations. This makes it much easier to integrate CompilerGym with your existing build scripts. - Added three new wrapper classes: `Counter`, that provides op counts for analysis ([#683](#683)); `SynchronousSqliteLogger`, that provides logging of environment interactions to a relational database ([#679](#679)), and `ForkOnStep` that provides an `undo()` operation ([#682](#682)). - Added `reward_space` and `observation_space` parameters to `env.reset()` ([#659](#659), thanks @SoumyajitKarmakar!) This release includes a number of improvements to the backend APIs that make it easier to write new CompilerGym environments: - Refactored the backend to make `CompilerEnv` an abstract interface, and `ClientServiceCompilerEnv` the concrete implementation of this interface. This enables new environments to be implemented without using gRPC ([#633](#633), thanks @sogartar!). - Extended the support for different types of action and observation spaces ([#641](#641), [#643](#643), thanks @sogartar!), including new `Permutation` and `SpaceSequence` spaces ([#645](#645), thanks @sogartar!).. - Added a new `disk/` subdirectory to compiler service's working directories, which is symlinked to an on-disk location for devices which support in-memory working directories. This fixes a bug with leftover temporary directories from LLVM ([#672](#672)). This release also includes numerous bug fixes and improvements, many of which were reported or fixed by the community. For example, fixing a bug in cache file locations ([#656](#656), thanks @uduse!), and a missing flag definition in example code ([#684](#684), thanks @anthony0727!). **Full Changelog**: v0.2.3...v0.2.4 This release brings in deprecating changes to the core `env.step()` routine, and lays the groundwork for enabling new types of compiler optimizations to be exposed through CompilerGym. Many thanks to code contributors: @mostafaelhoushi, @sogartar, @KyleHerndon, @uduse, @parthchadha, and @xtremey! Highlights of this release include: - Added a new `TextSizeInBytes` observation space for LLVM ([#575](#575)). * Added a new PPO leaderboard entry ([#580](#580). Thanks @xtremey! - Fixed a bug in which temporary directories created by the LLVM environment were not cleaned up ([#592](#592)). - **[Backend]** The function `createAndRunCompilerGymService` now returns an int, which is the exit return code ([#592](#592)). - Improvements to the examples documentation ([#548](#548)) and FAQ ([#586](#586)) Deprecations and breaking changes: - `CompilerEnv.step` no longer accepts a list of actions ([#627](#627)). A new method, `CompilerEnv.multistep` provides this functionality. This is to provide compatibility with environments whose action spaces are lists. To update your code, replace any calls to `env.step()` which take a list of actions to use `env.multistep()`. Thanks @sogartar! - The arguments `observations` and `rewards` to `step()` have been renamed `observation_spaces` and `reward_spaces`, respectively ([#627](#627)). - `Reward.id` has been renamed `Reward.name` ([#565](#565), [#612](#612)). Thanks @parthchadha! * The backend protocol buffer schema has been updated to natively support more types of observation and action, and to support nested spaces ([#531](#531)). Thanks @sogartar!
ChrisCummins
added a commit
that referenced
this pull request
May 24, 2022
This release adds a new compiler environment, new APIs, and a suite of backend improvements to improve the flexibility of CompilerGym environments. Many thanks to code contributors: @sogartar, @KyleHerndon, @SoumyajitKarmakar, @uduse, and @anthony0727! Highlights of this release include: - [mlir] Began work on a new environment for matrix multiplication using MLIR ([#652](#652), thanks @KyleHerndon and @sogartar!). Note this environment is not yet included in the pypi package and must be [compiled from source](https://github.com/facebookresearch/CompilerGym/blob/development/INSTALL.md#building-from-source-with-cmake). - [llvm] Added a new `env.benchmark_from_clang_invocation()` method ([#577](#577)) that can be used for constructing LLVM environment automatically from C/C++ compiler invocations. This makes it much easier to integrate CompilerGym with your existing build scripts. - Added three new wrapper classes: `Counter`, that provides op counts for analysis ([#683](#683)); `SynchronousSqliteLogger`, that provides logging of environment interactions to a relational database ([#679](#679)), and `ForkOnStep` that provides an `undo()` operation ([#682](#682)). - Added `reward_space` and `observation_space` parameters to `env.reset()` ([#659](#659), thanks @SoumyajitKarmakar!) This release includes a number of improvements to the backend APIs that make it easier to write new CompilerGym environments: - Refactored the backend to make `CompilerEnv` an abstract interface, and `ClientServiceCompilerEnv` the concrete implementation of this interface. This enables new environments to be implemented without using gRPC ([#633](#633), thanks @sogartar!). - Extended the support for different types of action and observation spaces ([#641](#641), [#643](#643), thanks @sogartar!), including new `Permutation` and `SpaceSequence` spaces ([#645](#645), thanks @sogartar!).. - Added a new `disk/` subdirectory to compiler service's working directories, which is symlinked to an on-disk location for devices which support in-memory working directories. This fixes a bug with leftover temporary directories from LLVM ([#672](#672)). This release also includes numerous bug fixes and improvements, many of which were reported or fixed by the community. For example, fixing a bug in cache file locations ([#656](#656), thanks @uduse!), and a missing flag definition in example code ([#684](#684), thanks @anthony0727!). **Full Changelog**: v0.2.3...v0.2.4 This release brings in deprecating changes to the core `env.step()` routine, and lays the groundwork for enabling new types of compiler optimizations to be exposed through CompilerGym. Many thanks to code contributors: @mostafaelhoushi, @sogartar, @KyleHerndon, @uduse, @parthchadha, and @xtremey! Highlights of this release include: - Added a new `TextSizeInBytes` observation space for LLVM ([#575](#575)). * Added a new PPO leaderboard entry ([#580](#580). Thanks @xtremey! - Fixed a bug in which temporary directories created by the LLVM environment were not cleaned up ([#592](#592)). - **[Backend]** The function `createAndRunCompilerGymService` now returns an int, which is the exit return code ([#592](#592)). - Improvements to the examples documentation ([#548](#548)) and FAQ ([#586](#586)) Deprecations and breaking changes: - `CompilerEnv.step` no longer accepts a list of actions ([#627](#627)). A new method, `CompilerEnv.multistep` provides this functionality. This is to provide compatibility with environments whose action spaces are lists. To update your code, replace any calls to `env.step()` which take a list of actions to use `env.multistep()`. Thanks @sogartar! - The arguments `observations` and `rewards` to `step()` have been renamed `observation_spaces` and `reward_spaces`, respectively ([#627](#627)). - `Reward.id` has been renamed `Reward.name` ([#565](#565), [#612](#612)). Thanks @parthchadha! * The backend protocol buffer schema has been updated to natively support more types of observation and action, and to support nested spaces ([#531](#531)). Thanks @sogartar!
ChrisCummins
added a commit
that referenced
this pull request
May 24, 2022
This release adds a new compiler environment, new APIs, and a suite of backend improvements to improve the flexibility of CompilerGym environments. Many thanks to code contributors: @sogartar, @KyleHerndon, @SoumyajitKarmakar, @uduse, and @anthony0727! Highlights of this release include: - [mlir] Began work on a new environment for matrix multiplication using MLIR ([#652](#652), thanks @KyleHerndon and @sogartar!). Note this environment is not yet included in the pypi package and must be [compiled from source](https://github.com/facebookresearch/CompilerGym/blob/development/INSTALL.md#building-from-source-with-cmake). - [llvm] Added a new `env.benchmark_from_clang_invocation()` method ([#577](#577)) that can be used for constructing LLVM environment automatically from C/C++ compiler invocations. This makes it much easier to integrate CompilerGym with your existing build scripts. - Added three new wrapper classes: `Counter`, that provides op counts for analysis ([#683](#683)); `SynchronousSqliteLogger`, that provides logging of environment interactions to a relational database ([#679](#679)), and `ForkOnStep` that provides an `undo()` operation ([#682](#682)). - Added `reward_space` and `observation_space` parameters to `env.reset()` ([#659](#659), thanks @SoumyajitKarmakar!) This release includes a number of improvements to the backend APIs that make it easier to write new CompilerGym environments: - Refactored the backend to make `CompilerEnv` an abstract interface, and `ClientServiceCompilerEnv` the concrete implementation of this interface. This enables new environments to be implemented without using gRPC ([#633](#633), thanks @sogartar!). - Extended the support for different types of action and observation spaces ([#641](#641), [#643](#643), thanks @sogartar!), including new `Permutation` and `SpaceSequence` spaces ([#645](#645), thanks @sogartar!).. - Added a new `disk/` subdirectory to compiler service's working directories, which is symlinked to an on-disk location for devices which support in-memory working directories. This fixes a bug with leftover temporary directories from LLVM ([#672](#672)). This release also includes numerous bug fixes and improvements, many of which were reported or fixed by the community. For example, fixing a bug in cache file locations ([#656](#656), thanks @uduse!), and a missing flag definition in example code ([#684](#684), thanks @anthony0727!). **Full Changelog**: v0.2.3...v0.2.4 This release brings in deprecating changes to the core `env.step()` routine, and lays the groundwork for enabling new types of compiler optimizations to be exposed through CompilerGym. Many thanks to code contributors: @mostafaelhoushi, @sogartar, @KyleHerndon, @uduse, @parthchadha, and @xtremey! Highlights of this release include: - Added a new `TextSizeInBytes` observation space for LLVM ([#575](#575)). * Added a new PPO leaderboard entry ([#580](#580). Thanks @xtremey! - Fixed a bug in which temporary directories created by the LLVM environment were not cleaned up ([#592](#592)). - **[Backend]** The function `createAndRunCompilerGymService` now returns an int, which is the exit return code ([#592](#592)). - Improvements to the examples documentation ([#548](#548)) and FAQ ([#586](#586)) Deprecations and breaking changes: - `CompilerEnv.step` no longer accepts a list of actions ([#627](#627)). A new method, `CompilerEnv.multistep` provides this functionality. This is to provide compatibility with environments whose action spaces are lists. To update your code, replace any calls to `env.step()` which take a list of actions to use `env.multistep()`. Thanks @sogartar! - The arguments `observations` and `rewards` to `step()` have been renamed `observation_spaces` and `reward_spaces`, respectively ([#627](#627)). - `Reward.id` has been renamed `Reward.name` ([#565](#565), [#612](#612)). Thanks @parthchadha! * The backend protocol buffer schema has been updated to natively support more types of observation and action, and to support nested spaces ([#531](#531)). Thanks @sogartar!
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This MR supersedes #611. It is the same thing except the source branch is different.