all AI news
Salesforce’s CodeRL Achieves SOTA Code Generation Results With Strong Zero-Shot Transfer Capabilities
Synced syncedreview.com
In the new paper CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning, a Salesforce Research team presents CodeRL, a novel framework for program synthesis tasks that employs pretrained language models (LMs) and deep reinforcement learning (RL) and achieves state-of-the-art performance on the challenging APPS benchmark while also demonstrating impressive zero-shot transfer capabilities.
The post Salesforce’s CodeRL Achieves SOTA Code Generation Results With Strong Zero-Shot Transfer Capabilities first appeared on Synced.
ai artificial intelligence code code generation deep-neural-networks generation machine learning machine learning & data science ml pretrained language model research salesforce sota technology transfer