Conversation
pjantrania
reviewed
Jun 14, 2024
bridger/go_explore_phase_1.py
Outdated
| def __init__(self, rng): | ||
| self._rng = rng | ||
|
|
||
| def update_times_since_lead_something_new(self, state, led_to_something_to_new: bool) -> None: |
Collaborator
There was a problem hiding this comment.
is “lead” different from “led” or “led to”?
Owner
Author
|
Definitely a draft pr and grammar issue o_O
________________________________
From: pjantrania ***@***.***>
Sent: Friday, June 14, 2024 7:58:42 AM
To: ldoshi/rome-wasnt-built-in-a-day ***@***.***>
Cc: Lyric P Doshi ***@***.***>; Author ***@***.***>
Subject: Re: [ldoshi/rome-wasnt-built-in-a-day] Initial impl of go explore. (PR #219)
@pjantrania commented on this pull request.
________________________________
In bridger/go_explore_phase_1.py<#219 (comment)>:
+ sampled_count: int = 0
+ visit_count: int = 1
+
***@***.***
+class SuccessEntry:
+ trajectory: list[int]
+ reward: float
+
+class StateCache:
+
+ _cache: dict[Any, CacheEntry] = {}
+
+ def __init__(self, rng):
+ self._rng = rng
+
+ def update_times_since_lead_something_new(self, state, led_to_something_to_new: bool) -> None:
is “lead” different from “led” or “led to”?
—
Reply to this email directly, view it on GitHub<#219 (review)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAUJWGBB2BX4GRDVRR26GSLZHMAKFAVCNFSM6AAAAABJJWPEK6VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDCMJYGYYDAMBXHA>.
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
Collaborator
|
thanks, I really was curious :)On Jun 14, 2024, at 11:34 AM, ldoshi ***@***.***> wrote:
Definitely a draft pr and grammar issue o_O
________________________________
From: pjantrania ***@***.***>
Sent: Friday, June 14, 2024 7:58:42 AM
To: ldoshi/rome-wasnt-built-in-a-day ***@***.***>
Cc: Lyric P Doshi ***@***.***>; Author ***@***.***>
Subject: Re: [ldoshi/rome-wasnt-built-in-a-day] Initial impl of go explore. (PR #219)
@pjantrania commented on this pull request.
________________________________
In bridger/go_explore_phase_1.py<#219 (comment)>:
+ sampled_count: int = 0
+ visit_count: int = 1
+
***@***.***
+class SuccessEntry:
+ trajectory: list[int]
+ reward: float
+
+class StateCache:
+
+ _cache: dict[Any, CacheEntry] = {}
+
+ def __init__(self, rng):
+ self._rng = rng
+
+ def update_times_since_lead_something_new(self, state, led_to_something_to_new: bool) -> None:
is “lead” different from “led” or “led to”?
—
Reply to this email directly, view it on GitHub<#219 (review)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAUJWGBB2BX4GRDVRR26GSLZHMAKFAVCNFSM6AAAAABJJWPEK6VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDCMJYGYYDAMBXHA>.
You are receiving this because you authored the thread.Message ID: ***@***.***>
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Owner
Author
|
:-) . This impl led to a assert in gym bridges that we never hit before. Fascinating
________________________________
From: pjantrania ***@***.***>
Sent: Friday, June 14, 2024 8:43:09 AM
To: ldoshi/rome-wasnt-built-in-a-day ***@***.***>
Cc: Lyric P Doshi ***@***.***>; Author ***@***.***>
Subject: Re: [ldoshi/rome-wasnt-built-in-a-day] Initial impl of go explore. (PR #219)
thanks, I really was curious :)On Jun 14, 2024, at 11:34 AM, ldoshi ***@***.***> wrote:
Definitely a draft pr and grammar issue o_O
________________________________
From: pjantrania ***@***.***>
Sent: Friday, June 14, 2024 7:58:42 AM
To: ldoshi/rome-wasnt-built-in-a-day ***@***.***>
Cc: Lyric P Doshi ***@***.***>; Author ***@***.***>
Subject: Re: [ldoshi/rome-wasnt-built-in-a-day] Initial impl of go explore. (PR #219)
@pjantrania commented on this pull request.
________________________________
In bridger/go_explore_phase_1.py<#219 (comment)>:
+ sampled_count: int = 0
+ visit_count: int = 1
+
***@***.***
+class SuccessEntry:
+ trajectory: list[int]
+ reward: float
+
+class StateCache:
+
+ _cache: dict[Any, CacheEntry] = {}
+
+ def __init__(self, rng):
+ self._rng = rng
+
+ def update_times_since_lead_something_new(self, state, led_to_something_to_new: bool) -> None:
is “lead” different from “led” or “led to”?
—
Reply to this email directly, view it on GitHub<#219 (review)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAUJWGBB2BX4GRDVRR26GSLZHMAKFAVCNFSM6AAAAABJJWPEK6VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDCMJYGYYDAMBXHA>.
You are receiving this because you authored the thread.Message ID: ***@***.***>
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>
—
Reply to this email directly, view it on GitHub<#219 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAUJWGHNGH2JHJ3I6OLZSQDZHMFQ3AVCNFSM6AAAAABJJWPEK6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNRYGI4TQMRYGU>.
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
…hing going backwards.
* Update go explore phase 1 to use multiprocessing * move success entry generation functionality to constructor
use cuda in a few places to enable use of gpu.
Support converting states to cells to allow various ideas for merging states together.
…the key is no longer a state (#235)
chunk start entries to send more work per process
add logging
In preparation for making the cache in rollout only hold diffs, remove some local vars and plumb hparams in deeper. still runs: python go_explore_phase_1.py --env-width=4 --go-explore-num-iterations=8 --cell-manager=downsample_cell_manager ========== Entry Count: 5 * wa-sampled: 0.1 * wa-new: 0 * wa-visit: 0.3 Shortest: [2, 2, 3, 3, 5]
Collaborator
|
Use |
* Added updated heatmap code * Fix issue with summing * rename to occurrenceentry
rollouts use cache diffs to track changes.
Collaborator
|
Let's add the |
Collaborator
|
Resolve circular import and get the data out of the cache_entry database. Should be easy to push that data to the frontend, where we want to see columns of data vertically arranged for the top n metrics sorted for cache entries |
make rollout workers run async relative to the cache updates in the main thread to increase throughput of rollouts.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
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.
No description provided.