4 ms·
RecurrentGemma: Moving Past Transformers for Efficient Open Language Models [pdf]
- aaboltaabol 2y agoCode here: https://github.com/google-deepmind/recurrentgemma https://github.com/google-deepmind/recurrentgemma Checkpoints here for both base pre-trained model and an IT version for dialogue: https://www.kaggle.com/models/google/recurrentgemma https://www.kaggle.com/models/google/recurrentgemma