Update README.md

main
PENG Bo 3 years ago committed by GitHub
parent 9721b8f9c5
commit f03efd0218
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

@ -1,8 +1,8 @@
# The RWKV Language Model (and my tricks for LMs) # The RWKV Language Model (and my LM tricks)
## RWKV: RNN with Transformer-level LLM Performance ## RWKV: RNN with Transformer-level LLM Performance
RWKV is a RNN with Transformer-level LLM performance, which can also be directly trained like a GPT transformer (parallelizable). And it's attention-free. You only need the hidden state at position t to compute the state at position t+1. You can use the "GPT" mode to quickly computer the hidden state for the "RNN" mode. RWKV is a RNN with Transformer-level LLM performance, which can also be directly trained like a GPT transformer (parallelizable). And it's 100% attention-free. You only need the hidden state at position t to compute the state at position t+1. You can use the "GPT" mode to quickly computer the hidden state for the "RNN" mode.
So it's combining the best of RNN and transformer - **great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding** (using the final hidden state). So it's combining the best of RNN and transformer - **great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding** (using the final hidden state).
@ -12,11 +12,11 @@ So it's combining the best of RNN and transformer - **great performance, fast in
![RWKV-chat](RWKV-chat.png) ![RWKV-chat](RWKV-chat.png)
**You can run RWKV-4 14B on any GPU using this fork (choose pytorch-stream):** https://github.com/harrisonvanderbyl/rwkv_chatbot **You can run RWKV on low VRAM GPUs with this fork (choose pytorch-stream):** https://github.com/harrisonvanderbyl/rwkv_chatbot
--- ---
I am training RWKV-4 14B on the Pile (final release around Jan-31-2023): https://wandb.ai/blinkdl/RWKV-v4-Pile I am training RWKV-4 14B on the Pile (final release around Feb-15-2023): https://wandb.ai/blinkdl/RWKV-v4-Pile
![RWKV-eval2](RWKV-eval2.png) ![RWKV-eval2](RWKV-eval2.png)

Loading…
Cancel
Save