| Challenge: | Recent work on pretrained language models for Hebrew is under-parameterized and under-trained . previous work on pretraining Hebrew LMs focused on encoder-only architectures . |
| Approach: | They propose to use sequence-to-sequence generative architectures to train large LMs in morphologically rich languages such as Hebrew. |
| Outcome: | The proposed model improves on all existing Hebrew NLP benchmarks. |
Similar Papers
AlephBERT: Language Model Pre-training and Evaluation from Sub-Word to Sentence Level (2022.acl-long)
Copied to clipboard
| Challenge: | a recent study shows that large pre-trained language models are not sufficient for Hebrew. |
| Approach: | They propose a large pre-trained language model for Hebrew that recovers morphological segments encoded in contextualized embedding vectors. |
| Outcome: | The proposed model obtains state-of-the-art on all tasks beyond contemporary Hebrew baselines. |
MRL Parsing Without Tears: The Case of Hebrew (2024.findings-acl)
Copied to clipboard
| Challenge: | a new approach to parsing morphologically rich languages (MRLs) is needed to overcome the deficiencies of current approaches. |
| Approach: | They propose a "flipped pipeline" where multiple layers are predicted independently on whole-token basis and then synthesized. |
| Outcome: | The proposed model achieves near-SOTA performance on Hebrew NLP tasks. |
Do Pretrained Contextual Language Models Distinguish between Hebrew Homograph Analyses? (2023.eacl-main)
Copied to clipboard
| Challenge: | Semitic morphologically-rich languages are characterized by extreme word ambiguity . many of the words are homographs with multiple possible analyses . |
| Approach: | They evaluate existing models for Hebrew homographs using word-piece embeddings . they find they are more effective when the number of word-part splits is limited . |
| Outcome: | The proposed models outperform non-contextualized embeddings on Hebrew homograph challenge sets. |
Combining Deep Generative Models and Multi-lingual Pretraining for Semi-supervised Document Classification (2021.eacl-main)
Copied to clipboard
| Challenge: | Semi-supervised learning and multilingual pretraining have been shown to be effective for task-specific labelled data shortages. |
| Approach: | They propose to combine semi-supervised deep generative models and multi-lingual pretraining to form a pipeline for document classification task. |
| Outcome: | The proposed method outperforms state-of-the-art models in low-resource settings across several languages and outperformed existing models in English. |
Sequence-to-Sequence Spanish Pre-trained Language Models (2024.lrec-main)
Copied to clipboard
| Challenge: | Spanish language models have demonstrated proficiency in natural language understanding and generation, but there is a scarcity of encoder-decoder models specifically designed for sequence-to-sequence tasks. |
| Approach: | They propose to implement encoder-decoder architectures pre-trained on Spanish corpora . they use them to assess sequence-to-sequence tasks including summarization, question answering . |
| Outcome: | The proposed models outperform models on sequence-to-sequence tasks in Spanish . the models show that they perform well across all tasks, the authors note . |
What is the best recipe for character-level encoder-only modelling? (2023.acl-long)
Copied to clipboard
| Challenge: | aims to benchmark recent progress in language understanding models that output contextualised representations at the character level. |
| Approach: | They aim to find the best way to build and train character-level BERT-like models by comparing architectural innovations with pretraining objectives. |
| Outcome: | The proposed model outperforms a token-based model on a set of evaluation tasks with a fixed training procedure. |
Multilingual Denoising Pre-training for Neural Machine Translation (2020.tacl-1)
Copied to clipboard
Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li, Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer
| Challenge: | Existing approaches to pre-train models focus on only English corpora, but this is not common in machine translation. |
| Approach: | They propose a sequence-to-sequence denoising auto-encoder pre-trained on monolingual corpora . they show that it produces significant performance gains across MT tasks . |
| Outcome: | The proposed model can achieve significant performance gains across a wide variety of MT tasks. |
nmT5 - Is parallel data still relevant for pre-training massively multilingual language models? (2021.acl-short)
Copied to clipboard
| Challenge: | Recent studies have shown that cross-lingual transfer learning in pre-trained multilingual models could be improved further by incorporating parallel data. |
| Approach: | They propose to integrate parallel data into mT5 pre-training to improve results on downstream multilingual and cross-lingual tasks. |
| Outcome: | The proposed model improves cross-lingual transfer significantly in small fine-tuning datasets and small model sizes. |
One Model is All You Need: ByT5-Sanskrit, a Unified Model for Sanskrit NLP Tasks (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Morphologically rich languages are notoriously challenging to process for downstream NLP applications. |
| Approach: | They propose a pretrained model for NLP applications involving the morphologically rich language Sanskrit that outperforms previous models by a considerable margin. |
| Outcome: | The proposed model outperforms tokenized models on established Sanskrit word segmentation tasks and matches the current best lexicon-based model. |
What’s Wrong with Hebrew NLP? And How to Make it Right (D19-3)
Copied to clipboard
| Challenge: | Sub-optimal performance of many morphologically rich languages (MRLs) is due to errors in early morphology disambiguation decisions, that cannot be recovered later on in the pipeline, yielding incoherent annotations on the whole. |
| Approach: | They propose to use a joint morpho-syntactic infrastructure for processing Modern Hebrew texts to provide rich and expressive annotations. |
| Outcome: | The proposed pipelines are based on a morpho-syntactic infrastructure for processing Modern Hebrew texts. |