Feat: Add support for parquet files

[deependujha]

Before submitting

• Was this discussed/agreed via a Github issue? (no need for typos and docs improvements)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure to update the docs?
• Did you write any new necessary tests?

What does this PR do?

Fixes #191

• Index parquet dataset stored in local or cloud (s3 or gs).

import litdata as ld

pq_data_uri = "gs://deep-litdata-parquet/my-parquet-data"

ld.index_parquet_dataset(pq_data_uri)

• Use it as normal optimized dataset

import litdata as ld
from litdata.streaming.item_loader import ParquetLoader

ds = ld.StreamingDataset('gs://deep-litdata-parquet/my-parquet-data', item_loader = ParquetLoader())

for _ds in ds:
    print(f"{_ds=}")

---

Benchmark on Data prep machine

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in GitHub issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃

[codecov]

Codecov Report

Attention: Patch coverage is 64.70588% with 54 lines in your changes missing coverage. Please review.

Project coverage is 78%. Comparing base (ee77852) to head (59d9d14).
Report is 1 commits behind head on main.

Additional details and impacted files

@@         Coverage Diff          @@
##           main   #443    +/-   ##
====================================
- Coverage    78%    78%    -0%     
====================================
  Files        36     37     +1     
  Lines      5217   5365   +148     
====================================
+ Hits       4088   4182    +94     
- Misses     1129   1183    +54     

[tchaton]

Hey @deependujha Nice progress ;)

[tchaton]

This is dope. If we could automatically index an s3 folder

And generate an index file, it would be dope.

import polars as pl
import fsspec

file_path = "s3://your-bucket/path/to/your-file.parquet"

# Open the Parquet file with fsspec
with fsspec.open(file_path, mode="rb") as f:
    # Fetch the number of rows from the metadata
    num_rows = pl.read_parquet(f, use_pyarrow=True).shape[0]
    print(f"Number of rows: {num_rows}")

[gitguardian]

⚠️ GitGuardian has uncovered 1 secret following the scan of your pull request.

Please consider investigating the findings and remediating the incidents. Failure to do so may lead to compromising the associated services or software components.

Since your pull request originates from a forked repository, GitGuardian is not able to associate the secrets uncovered with secret incidents on your GitGuardian dashboard.
Skipping this check run and merging your pull request will create secret incidents on your GitGuardian dashboard.

🔎 Detected hardcoded secret in your pull request

GitGuardian id	GitGuardian status	Secret	Commit	Filename
5685611	Triggered	Generic High Entropy Secret	76efafb	tests/streaming/test_resolver.py	View secret

🛠 Guidelines to remediate hardcoded secrets

1. Understand the implications of revoking this secret by investigating where it is used in your code.
2. Replace and store your secret safely. Learn here the best practices.
3. Revoke and rotate this secret.
4. If possible, rewrite git history. Rewriting git history is not a trivial act. You might completely break other contributing developers' workflow and you risk accidentally deleting legitimate data.

To avoid such incidents in the future consider

• following these best practices for managing and storing secrets including API keys and other credentials
• install secret detection on pre-commit to catch secret before it leaves your machine and ease remediation.

---

^{🦉 GitGuardian detects secrets in your source code to help developers and security teams secure the modern development process. You are seeing this because you or someone else with access to this repository has authorized GitGuardian to scan your pull request.}

[deependujha]

Adding support for directly consuming HF datasets is an exciting direction!

For HF datasets, my current idea involves iterating through all the Parquet datasets in the HF repository and creating an index.json file that is stored in a cache (since modifying the original dataset is not feasible).

When using the streaming dataset/dataloader, we would then pass this separate index.json file from the cache.

At this point, I'm uncertain about the exact approach for handling HF datasets comprehensively. This PR is ready for review and lays the groundwork for future enhancements. We can discuss HF dataset integration in a subsequent PR.

[tchaton]

This is quite awesome !

[tchaton]

Can you add the benchmarks in the description ?

[justusschock]

looks great!

[lantiga]

Looks great Deepend

[lantiga]

"Index" may not be immediately clear to users imo.
Ultimately what users get is the ability to "Stream Parquet datasets", I'd have this as the title. Index is a technical detail.

[lantiga]

I'd also add a line or two explaining how big of a deal this is : ) "Stream Parquet files directly without converting them to the LitData optimized binary format" or something of this nature.

[deependujha]

I've the changes. Since this PR was already merged, so the new changes are in: PR: #460

[lantiga]

Might be good to prepend "You are using the Parquet item loader, which depends on Polars. Please run: pip install polars"

[lantiga]

Do we need to make bound checks on the version?

[deependujha]

I'm not sure on the exact version bound. To be on safer side for now, I've simply updated to polars>1.0.0.

Is this fine, or should I refine it further?

[lantiga]

Is there a fundamental reason why we're not pre-loading or is it just for sequencing?

[deependujha]

Apologies for the oversight. Thanks for pointing that out!

I've made the necessary changes now to include pre-loading as suggested.