Fengshenbang Scientific and Technological Achievements
Fengshenbang 1.0 : Fengshenbang Open Source Project 1.0 Chinese-English bilingual general paper, aims to become the infrastructure of Chinese cognitive intelligence.
BioBART : A generative language model in the biomedical field provided by Tsinghua University and IDEA Research Institute. (
BioNLP 2022)
> UniMC : A unified model for label-based datasets in zero-shot scenarios. ( EMNLP 2022 )
> FMIT : A single-tower multimodal named entity recognition model based on relative position encoding. ( COLING 2022 )
> UniEX : A Natural Language Understanding Model for Unified Extraction Tasks. ( ACL 2023 )
> Solving Math Word Problems via Cooperative Reasoning induced Language Models : Solving mathematical problems using a collaborative reasoning framework of language models. ( ACL 2023 )
> MVP-Tuning : A Parameter-Efficient Commonsense Question Answering System Based on Multi-View Knowledge Retrieval. ( ACL 2023 )
The big event of Fengshenbang
- Multimodal Ziya is online! Jiang Ziya General Model Vertical Capability Series Vol.1 Released 2023.06.05
- IDEA Research Institute’s Fengshenbang team launched another open source general model series "Jiang Ziya" 2023.05.17 -The first Chinese Stable Diffusion model open source , IDEA Research Institute Fengshenbang team opened the era of Chinese AI art 2022.11.2 -Breaking the impossible triangle, comparable to 540 billion models, IDEA Fengshenbang team only reached zero-sample learning SOTA for 200 million-level models 2022.10.25
- The winner of the AIWIN competition, Fengshenbang proposed a multi-task learning solution Uber 2022.07.21 -Finetune, the "Fengshenbang" pre-trained language model "Erlangshen" won the SimCLUE list 1 2022.07.14 -Fengshen framework is officially open source, Help you easily pre-train and fine-tune the major models of "Fengshenbang" 2022.06.30 -GTS model production platform is open for public beta, using AI to automate the production of AI models2022.05.23 - dataset released! IDEA Research Institute CCNL×NLPCC 2022 Task Challenge has begun, and the winning team will get the IDEA internship opportunity 2022.04.07 -Refreshed again! IDEA CCNL pre-trained language model "Erlangshen", this time won ZeroCLUE 2022.01.24
- IDEA Friends \| CCNL Team "Fengshen List", why did they choose IDEA? 2022.01.12
- IDEA conference release｜"Fengshen List" large-scale model open source plan 2021.11.25
- IDEA Research Institute Chinese pre-training model Erlang Shen topped the FewCLUE list 2021.11.11
Visit Official Website
