google-research-datasets / recognizing-multimodal-entailment
The dataset consists of public social media url pairs and the corresponding entailment label for an external conference (ACL 2021). Each url contains a post with both linguistic (text) and visual (image) content. Entailment labels are human annotated through Google Crowdsource.
☆13Updated 3 years ago
Related projects ⓘ
Alternatives and complementary repositories for recognizing-multimodal-entailment
- DEMix Layers for Modular Language Modeling☆53Updated 3 years ago
- ☆36Updated 7 months ago
- [ICML 2023] Code for our paper “Compositional Exemplars for In-context Learning”.☆92Updated last year
- Pytorch version of VidLanKD: Improving Language Understanding viaVideo-Distilled Knowledge Transfer (NeurIPS 2021))☆56Updated last year
- VaLM: Visually-augmented Language Modeling. ICLR 2023.☆56Updated last year
- [ICML 2023] Exploring the Benefits of Training Expert Language Models over Instruction Tuning☆97Updated last year
- TBC☆26Updated 2 years ago
- An Empirical Study On Contrastive Search And Contrastive Decoding For Open-ended Text Generation☆26Updated 5 months ago
- IntructIR, a novel benchmark specifically designed to evaluate the instruction following ability in information retrieval models. Our foc…☆28Updated 4 months ago
- ☆13Updated 2 years ago
- PyTorch reimplementation of REALM and ORQA☆22Updated 2 years ago
- This is the oficial repository for "Parameter-Efficient Multi-task Tuning via Attentional Mixtures of Soft Prompts" (EMNLP 2022)☆97Updated last year
- NAACL 2022: MCSE: Multimodal Contrastive Learning of Sentence Embeddings☆53Updated 5 months ago
- ☆76Updated last year
- [EMNLP 2022] Code for our paper “ZeroGen: Efficient Zero-shot Learning via Dataset Generation”.☆46Updated 2 years ago
- Dataset & Code for Com2Sense Benchmark☆13Updated 3 years ago
- [NeurIPS 2022] Generating Training Data with Language Models: Towards Zero-Shot Language Understanding☆63Updated 2 years ago
- Adding new tasks to T0 without catastrophic forgetting☆30Updated 2 years ago
- Staged Training for Transformer Language Models☆30Updated 2 years ago
- The code for lifelong few-shot language learning☆53Updated 2 years ago
- ☆26Updated last year
- ACL22 paper: Imputing Out-of-Vocabulary Embeddings with LOVE Makes Language Models Robust with Little Cost☆39Updated 11 months ago
- Code for paper "CrossFit : A Few-shot Learning Challenge for Cross-task Generalization in NLP" (https://arxiv.org/abs/2104.08835)☆104Updated 2 years ago
- Code and data for paper "Context-faithful Prompting for Large Language Models".☆39Updated last year
- A categorical archive of ChatGPT failures☆64Updated last year
- This repository accompanies our paper “Do Prompt-Based Models Really Understand the Meaning of Their Prompts?”☆84Updated 2 years ago
- ☆13Updated 11 months ago
- Supporting code for ReCEval paper☆26Updated last month
- ☆14Updated 2 years ago
- ☆14Updated last year