Fix outdated verification_mode values
#5607
Merged
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
|
The documentation is not available anymore as the PR was closed or merged. |
…erna/datasets into do-not-save-info-when-no-checks
…hange values again
lhoestq
reviewed
Mar 8, 2023
src/datasets/builder.py
Outdated
Comment on lines
360
to
361
| if os.path.exists(self._cache_dir) > 0: # check if data exist | ||
| if len(os.listdir(self._cache_dir)): |
There was a problem hiding this comment.
Suggested change
| if os.path.exists(self._cache_dir) > 0: # check if data exist | |
| if len(os.listdir(self._cache_dir)): | |
| if os.path.exists(self._cache_dir): # check if data exist | |
| if len(os.listdir(self._cache_dir)) > 0: |
src/datasets/builder.py
Outdated
Comment on lines
884
to
885
| if verification_mode is not VerificationMode.NO_CHECKS: | ||
| self._save_info() |
There was a problem hiding this comment.
We should always save the info. For example it stores the feature types and the arrow shards lengths.
There was a problem hiding this comment.
hm true...
what should we do then? I don't think it's fine that we get different results running the same command, like:
load_dataset("ds_name") # error - failed verifications
# then
load_dataset("ds_name", verification_mode="no_checks") # successfully loaded
# then
load_dataset("ds_name") # successfully loaded despite it was an error befoream I missing something?
There was a problem hiding this comment.
I don't think we can do anything tbh - e.g. reloading a cached dataset works even if you don't have the downloaded raw files anymore so we can't check for their checksums
There was a problem hiding this comment.
okay reverted it.
But I also changed incorrect values of verification_mode in two places, from "none" -> "no_checks", "all" -> "all_checks" (they weren't updated in the previous verification PR #5303), do you agree with this change?
polinaeterna
changed the title
verification_mode values
lhoestq
reviewed
Mar 8, 2023
| if len(os.listdir(self._cache_dir)) > 0 and os.path.exists( | ||
| path_join(self._cache_dir, config.DATASET_INFO_FILENAME) | ||
| ): | ||
| logger.info("Overwrite dataset info from restored data version if exists.") | ||
| self.info = DatasetInfo.from_directory(self._cache_dir) | ||
| else: # dir exists but no data, remove the empty dir as data aren't available anymore |
There was a problem hiding this comment.
I think we should remove the directory only if it's completely empty - but with your change it would remove it if the dataset info file is not there. This is in case some data fiels are still left somehow
There was a problem hiding this comment.
thanks for noticing, fixed it, can I merge?
Show benchmarksPyArrow==8.0.0 Show updated benchmarks!Benchmark: benchmark_array_xd.json
Benchmark: benchmark_getitem_100B.json
Benchmark: benchmark_indices_mapping.json
Benchmark: benchmark_iterating.json
Benchmark: benchmark_map_filter.json
Show updated benchmarks!Benchmark: benchmark_array_xd.json
Benchmark: benchmark_getitem_100B.json
Benchmark: benchmark_indices_mapping.json
Benchmark: benchmark_iterating.json
Benchmark: benchmark_map_filter.json
|
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.
I think it makes sense not to savedataset_info.jsonfile to a dataset cache directory when loading dataset withverification_mode="no_checks"because otherwise when next time the dataset is loaded withoutverification_mode="no_checks", it will be loaded successfully, despite some values in info might not correspond to the ones in the repo which was the reason for usingverification_mode="no_checks"first.Updated values of
verification_modeto the current ones in some places ("none" -> "no_checks", "all" -> "all_checks")