𝐂𝐨𝐝𝐞𝐆𝐞𝐦𝐦𝐚 : 𝐂𝐨𝐝𝐢𝐧𝐠 𝐋𝐋𝐌 𝐛𝐚𝐬𝐞𝐝 𝐨𝐧 𝐆𝐨𝐨𝐠𝐥𝐞’𝐬 𝐆𝐞𝐦𝐦𝐚 CodeGemma, a collection of open code models based on Google DeepMind’s Gemma models. Continuing from Gemma pretrained models, CodeGemma models are further trained on more than 500 to 1000 billion tokens of primarily code, using the same architectures as the Gemma model family. CodeGemma 2B is a state-of-the-art code completion model designed for fast code infilling and open-ended generation in latency-sensitive settings. CodeGemma paper details (in the comment) #llms #generativeai #nlproc #deeplearning #coding #codellms
Exciting development! 🌟 CodeGemma leverages Google DeepMind’s Gemma models, enhancing them with training on vast code datasets. CodeGemma 2B stands out for its state-of-the-art capabilities in code completion and generation, ideal for latency-sensitive applications. #CodeGemma #DeepLearning #AI #CodeCompletion #TechInnovation 🖥️🚀
I'll keep this in mind
CodeGemma paper - https://arxiv.org/abs/2406.11409