🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.
Repository Index
47,133 repositories indexed
1–4 of 4 · Most stars
Ongoing research training transformer models at scale
RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.
A PyTorch-based Speech Toolkit