lhoestq Follow

lhoestq Follow

Quentin Lhoest lhoestq

ML Engineer @huggingface | Maintainer of 🤗Datasets

Follow

47 followers · 17 following · 10

Hugging Face
Paris

Highlights

Arctic Code Vault Contributor

Organizations

Pinned

huggingface/datasets

🤗 Fast, efficient, open-access datasets and evaluation metrics in PyTorch, TensorFlow, NumPy and Pandas

Python 4.9k 382
huggingface/transformers

🤗Transformers: State-of-the-art Natural Language Processing for Pytorch and TensorFlow 2.0.

Python 37.3k 9.1k

1,038 contributions in the last year

Activity overview

Contributed to huggingface/datasets, huggingface/transformers, eltoto1219/nlp and 2 other repositories

Contribution activity

November 2020

Created 27 commits in 2 repositories

Created a pull request in huggingface/transformers that received 20 comments

Fix rag finetuning + add finetuning test

Following #7715 we need more test coverage of the RAG example scripts. In this PR I'm adding a test for the finetuning script. The test includes a …

+200 −56 • 20 comments

Opened 24 other pull requests in 2 repositories

huggingface/datasets

17 merged 3 open 1 closed

Fix url with backslash in windows for blimp and pg19 Nov 27
Disallow backslash in urls Nov 27
Follow cache_dir parameter to gcs downloader Nov 27
Allow arrow based builder in auto dummy data generation Nov 27
Better messages regarding split naming Nov 26
Allow several tags sets Nov 26
Add a few datasets of reference in the documentation Nov 26
Fix wikipedia custom config Nov 25
Auto generate dummy data Nov 24
Use GCP download url instead of tensorflow custom download for boolq Nov 24
Update ner datasets infos Nov 19
Consistent metric outputs Nov 18
Fix some metrics feature types Nov 18
Add clear_cache parameter in the test command Nov 17
Update head requests Nov 17
Integrate file_lock inside the lib for better logging control Nov 16
Use pandas reader in csv Nov 16
Fix kor nli csv reader Nov 16
Add writer_batch_size attribute to GeneratorBasedBuilder Nov 10
Make save function use deterministic global vars order Nov 9
Fix type hints pickling in python 3.6 Nov 9

huggingface/transformers

2 merged 1 closed

Reviewed 28 pull requests in 2 repositories

huggingface/datasets 23 pull requests

Fix URL with backslash in Windows Nov 27
Optional per-dataset default config name Nov 27
Addition of Nl2Bash Dataset Nov 27
add metrec: arabic poetry dataset Nov 27
Add template and documentation for dataset card Nov 27
Add LER Nov 26
Add SemEval-2010 task 8 Nov 26
gitignore .python-version Nov 26
Add IndicGLUE dataset and Metrics Nov 25
Auto generate dummy data Nov 24
Add open book corpus Nov 17
Create ClassLabel for labelling tasks datasets Nov 16
CNN/Dailymail Dataset Card Nov 12
AlloCiné dataset card Nov 12
Update loading_metrics.rst Nov 11
Update squad_v2.py Nov 11
Adding nli_tr dataset Nov 11
Add accuracy, precision, recall and F1 metrics Nov 10
dataset(ncslgr): add initial loading script Nov 10
Add ASNQ dataset Nov 10
fix: typos in tutorial to map KILT and TriviaQA Nov 10
Update quail dataset to v1.3 Nov 10
Fix seqeval metric Nov 6

huggingface/transformers 5 pull requests

Fix dpr<>bart config for RAG Nov 26
2 typos in modeling_rag.py Nov 25
Fix run_ner script Nov 19
Tokenizers: ability to load from model subfolder Nov 17
Add TFDPR Nov 11

Created an issue in huggingface/datasets that received 1 comment

[Caching] Dill globalvars() output order is not deterministic and can cause cache issues.

Dill uses dill.detect.globalvars to get the globals used by a function in a recursive dump. globalvars returns a dictionary of all the globals that…

1 comment

You can’t perform that action at this time.