DaSciM

university

https://www.lix.polytechnique.fr/dascim/

dascim_polytech

Activity Feed

AI & ML interests

Machine and deep learning for graphs, NLP, LLMs, Foundation Models for Biology

Recent Activity

mkonomi updated a dataset about 2 months ago

dascim/GreekMMLU

mkonomi authored a paper about 2 months ago

GreekMMLU: A Native-Sourced Multitask Benchmark for Evaluating Language Models in Greek

mkonomi updated a dataset about 2 months ago

dascim/GreekMMLU

View all activity

dascim 's collections 7

Prot2Text-V2

Running on Zero

1

Prot2Text-V2

🧬

1

Protein
xiao-fei/Prot2Text-V2-11B-Instruct-hf

Text Generation • 11B • Updated Oct 23, 2025 • 2.62k • 3
Prot2Text-V2: Protein Function Prediction with Multimodal Contrastive Alignment

Paper • 2505.11194 • Published May 16, 2025 • 2

BARThez

French sequence to sequence pretrained models based on BART.

BARThez: a Skilled Pretrained French Sequence-to-Sequence Model

Paper • 2010.12321 • Published Oct 23, 2020 • 1
moussaKam/barthez

Fill-Mask • Updated Feb 20, 2024 • 687 • 19
moussaKam/mbarthez

Fill-Mask • Updated Feb 20, 2024 • 117k • • 7
moussaKam/barthez-orangesum-abstract

Summarization • Updated Nov 15, 2021 • 698 • 7

FrugalScore

FrugalScore is a reference-based metric for NLG models evaluation.

FrugalScore: Learning Cheaper, Lighter and Faster Evaluation Metricsfor Automatic Text Generation

Paper • 2110.08559 • Published Oct 16, 2021
moussaKam/frugalscore_tiny_bert-base_bert-score

Text Classification • Updated Feb 1, 2022 • 2.49k
moussaKam/frugalscore_medium_bert-base_mover-score

Text Classification • Updated May 11, 2022 • 15
moussaKam/frugalscore_medium_roberta_bert-score

Text Classification • Updated Feb 1, 2022 • 164

GreekBART

Greek sequence to sequence models based on BART

GreekBART: The First Pretrained Greek Sequence-to-Sequence Model

Paper • 2304.00869 • Published Apr 3, 2023
dascim/greekbart

Fill-Mask • 0.1B • Updated Oct 15, 2024 • 11
dascim/greekbart-news24-abstract

Text Generation • 0.1B • Updated Oct 15, 2024 • 20
dascim/greekbart-news24-title

Text Generation • 0.1B • Updated Oct 15, 2024 • 17

Prot2Text

A suite of models to generate protein's function.

Running

Prot2Text

🧬

Chatbot
Sleeping

ESM2Text

🧬

Chatbot
Prot2Text: Multimodal Protein's Function Generation with GNNs and Transformers

Paper • 2307.14367 • Published Jul 25, 2023 • 3
habdine/Prot2Text-Base-v1-1

Text Generation • 0.3B • Updated Oct 23, 2025 • 17 • 4

BERTweetFR

BERTweetFR is the first large-scale pre-trained language model for French tweets.

Yanzhu/bertweetfr-base

Fill-Mask • Updated Jun 13, 2021 • 155 • 5
Yanzhu/bertweetfr_ner

Token Classification • Updated Sep 29, 2021 • 1
Yanzhu/bertweetfr_offensiveness

Updated Feb 4, 2022

JuriBERT

Set of BERT models pre-trained from scratch on French legal-domain text

JuriBERT: A Masked-Language Model Adaptation for French Legal Text

Paper • 2110.01485 • Published Oct 4, 2021
dascim/juribert-base

Fill-Mask • 0.1B • Updated Oct 14, 2024 • 390k
dascim/juribert-small

Fill-Mask • 42.2M • Updated Oct 14, 2024 • 1 • 1
dascim/juribert-mini

Fill-Mask • 15.8M • Updated Oct 14, 2024 • 3

Prot2Text-V2

Running on Zero

1

Prot2Text-V2

🧬

1

Protein
xiao-fei/Prot2Text-V2-11B-Instruct-hf

Text Generation • 11B • Updated Oct 23, 2025 • 2.62k • 3
Prot2Text-V2: Protein Function Prediction with Multimodal Contrastive Alignment

Paper • 2505.11194 • Published May 16, 2025 • 2

Prot2Text

A suite of models to generate protein's function.

Running

Prot2Text

🧬

Chatbot
Sleeping

ESM2Text

🧬

Chatbot
Prot2Text: Multimodal Protein's Function Generation with GNNs and Transformers

Paper • 2307.14367 • Published Jul 25, 2023 • 3
habdine/Prot2Text-Base-v1-1

Text Generation • 0.3B • Updated Oct 23, 2025 • 17 • 4

BARThez

French sequence to sequence pretrained models based on BART.

BARThez: a Skilled Pretrained French Sequence-to-Sequence Model

Paper • 2010.12321 • Published Oct 23, 2020 • 1
moussaKam/barthez

Fill-Mask • Updated Feb 20, 2024 • 687 • 19
moussaKam/mbarthez

Fill-Mask • Updated Feb 20, 2024 • 117k • • 7
moussaKam/barthez-orangesum-abstract

Summarization • Updated Nov 15, 2021 • 698 • 7

BERTweetFR

BERTweetFR is the first large-scale pre-trained language model for French tweets.

Yanzhu/bertweetfr-base

Fill-Mask • Updated Jun 13, 2021 • 155 • 5
Yanzhu/bertweetfr_ner

Token Classification • Updated Sep 29, 2021 • 1
Yanzhu/bertweetfr_offensiveness

Updated Feb 4, 2022

FrugalScore

FrugalScore is a reference-based metric for NLG models evaluation.

FrugalScore: Learning Cheaper, Lighter and Faster Evaluation Metricsfor Automatic Text Generation

Paper • 2110.08559 • Published Oct 16, 2021
moussaKam/frugalscore_tiny_bert-base_bert-score

Text Classification • Updated Feb 1, 2022 • 2.49k
moussaKam/frugalscore_medium_bert-base_mover-score

Text Classification • Updated May 11, 2022 • 15
moussaKam/frugalscore_medium_roberta_bert-score

Text Classification • Updated Feb 1, 2022 • 164

JuriBERT

Set of BERT models pre-trained from scratch on French legal-domain text

JuriBERT: A Masked-Language Model Adaptation for French Legal Text

Paper • 2110.01485 • Published Oct 4, 2021
dascim/juribert-base

Fill-Mask • 0.1B • Updated Oct 14, 2024 • 390k
dascim/juribert-small

Fill-Mask • 42.2M • Updated Oct 14, 2024 • 1 • 1
dascim/juribert-mini

Fill-Mask • 15.8M • Updated Oct 14, 2024 • 3

GreekBART

Greek sequence to sequence models based on BART

GreekBART: The First Pretrained Greek Sequence-to-Sequence Model

Paper • 2304.00869 • Published Apr 3, 2023
dascim/greekbart

Fill-Mask • 0.1B • Updated Oct 15, 2024 • 11
dascim/greekbart-news24-abstract

Text Generation • 0.1B • Updated Oct 15, 2024 • 20
dascim/greekbart-news24-title

Text Generation • 0.1B • Updated Oct 15, 2024 • 17

AI & ML interests

Recent Activity

Team members 7

dascim 's collections 7

Prot2Text-V2

Prot2Text

ESM2Text

Prot2Text-V2

Prot2Text

ESM2Text