Skip to content

Commit

Permalink
Update README.md
Browse files Browse the repository at this point in the history
  • Loading branch information
BlinkDL authored Dec 16, 2023
1 parent 5fc3ff9 commit 9eabe76
Showing 1 changed file with 6 additions and 0 deletions.
6 changes: 6 additions & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,12 @@ cd RWKV-v5/
./demo-training-run.sh
(you may want to log in to wandb first)
```
Your loss curve should look almost exactly the same as this, with the same ups and downs (if you use the same bsz & config):

![RWKV-v5-minipile](RWKV-v5-minipile.png)

You can run your model using https://pypi.org/project/rwkv/ (use "rwkv_vocab_v20230424" instead of "20B_tokenizer.json")

## RWKV: Parallelizable RNN with Transformer-level LLM Performance (pronounced as "RwaKuv", from 4 major params: R W K V)

RWKV is an 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 compute the hidden state for the "RNN" mode.
Expand Down

0 comments on commit 9eabe76

Please sign in to comment.